WO2023016085A1 - 一种业务数据处理方法、装置、电子设备、计算机可读存储介质及计算机程序产品 - Google Patents

一种业务数据处理方法、装置、电子设备、计算机可读存储介质及计算机程序产品 Download PDF

Info

Publication number
WO2023016085A1
WO2023016085A1 PCT/CN2022/099533 CN2022099533W WO2023016085A1 WO 2023016085 A1 WO2023016085 A1 WO 2023016085A1 CN 2022099533 W CN2022099533 W CN 2022099533W WO 2023016085 A1 WO2023016085 A1 WO 2023016085A1
Authority
WO
WIPO (PCT)
Prior art keywords
business
data
unconverted
target
business data
Prior art date
Application number
PCT/CN2022/099533
Other languages
English (en)
French (fr)
Inventor
李邦鹏
Original Assignee
腾讯科技（深圳）有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 腾讯科技（深圳）有限公司 filed Critical 腾讯科技（深圳）有限公司
Publication of WO2023016085A1 publication Critical patent/WO2023016085A1/zh
Priority to US18/316,081 priority Critical patent/US20230281662A1/en

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce
    • G06Q30/02—Marketing; Price estimation or determination; Fundraising
    • G06Q30/0241—Advertisements
    • G06Q30/0277—Online advertisement
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce
    • G06Q30/02—Marketing; Price estimation or determination; Fundraising
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce
    • G06Q30/02—Marketing; Price estimation or determination; Fundraising
    • G06Q30/0241—Advertisements
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce
    • G06Q30/02—Marketing; Price estimation or determination; Fundraising
    • G06Q30/0241—Advertisements
    • G06Q30/0251—Targeted advertisements
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce
    • G06Q30/02—Marketing; Price estimation or determination; Fundraising
    • G06Q30/0241—Advertisements
    • G06Q30/0251—Targeted advertisements
    • G06Q30/0254—Targeted advertisements based on statistics

Definitions

  • the present application relates to the field of computer technology, and in particular to a business data processing method, device, electronic equipment, computer readable storage medium and computer program product.
  • the application platform will first sort the various business data, and then deliver the sorted business data to the user groups in sequence.
  • the conversion rate that is, the ratio between the number of users who place orders and purchases through business data and the number of users who click on business data
  • related technologies will be repeated many times and always in the same order. All business data is delivered.
  • this method will cause excessive delivery of business data and cause a large waste of platform resources (computing resources such as processor threads, communication resources such as bandwidth). Data processing costs, and it is difficult to improve the accuracy of business data delivery.
  • Embodiments of the present application provide a service data processing method, device, electronic equipment, computer-readable storage medium, and computer program product, which can improve the accuracy of service data delivery, improve the accuracy of service data delivery, and save resources.
  • the embodiment of this application provides a business data processing method, including:
  • the delivery log includes the business output status and business conversion status of the business object set respectively for the N delivered business data;
  • N is a positive integer;
  • the business conversion status includes the unconverted status;
  • the business output state includes the output state;
  • the business output status and business conversion status of N delivered business data respectively according to the business object set, establish the untransformed browsing behavior characteristics of the target business object for the unconverted business data;
  • the business output status of the target business object for the untransformed business data is The status has been output and the status of business conversion is unconverted;
  • the N delivered business data includes unconverted business data, and the set of business objects includes the target business object;
  • unconverted browsing behavior determine the sorting factor corresponding to the target business object and the unconverted item, sort the business data containing the unconverted item according to the sorting factor corresponding to the target business object and the unconverted item, and obtain the sequence business Data, select the target business data from the sequence business data according to the order of sorting and deliver it to the target business object; unconverted business data includes unconverted items.
  • An embodiment of the present application provides a business data processing device, including:
  • the log acquisition module is used to obtain the delivery log corresponding to the N delivered business data; the delivery log includes the business output status and the business conversion status of the business object set respectively for the N delivered business data; N is a positive integer; the business conversion status includes Unconverted status; business output status includes output status;
  • the feature building module is used to establish the unconverted browsing behavior characteristics of the target business object for the unconverted business data according to the business output status and business conversion status of the N delivered business data according to the business object set;
  • the target business object is for the unconverted business
  • the business output status of the data is the output status and the business conversion status is the unconverted status;
  • the N delivered business data includes untransformed business data, and the business object set includes the target business object;
  • the sorting factor determination module is used to determine the sorting factor corresponding to the target business object and the unconverted item according to the characteristics of the unconverted browsing behavior; the unconverted business data includes the unconverted item.
  • the data sorting module is used to sort the business data containing the unconverted items according to the sorting factor corresponding to the target business object and the unconverted items, and obtain the sequence business data;
  • the data delivery module is used to select target business data from the sequence business data in order of sorting and deliver it to the target business object.
  • An embodiment of the present application provides an electronic device, including: a processor and a memory;
  • the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the method in the embodiment of the present application.
  • An embodiment of the present application provides a computer-readable storage medium, where a computer program is stored in the computer-readable storage medium, and the computer program includes program instructions.
  • the program instructions are executed by a processor, the method in the embodiment of the present application is executed.
  • An embodiment of the present application provides a computer program product or computer program, the computer program product or computer program includes computer instructions, the computer instructions are stored in a computer-readable storage medium, and the processor of the electronic device reads the computer-readable storage medium from the computer-readable storage medium.
  • the computer instruction is fetched, and the processor executes the computer instruction, so that the electronic device executes the method provided in the embodiment of the present application.
  • the target business object since the target business object’s business output status for untransformed business data is output status, it indicates that the target business object is interested in the business items in the untransformed business data; thus, by calculating the target business object The sorting factor corresponding to the business item, and then sort the business data containing the same unconverted item according to the sorting factor, and select the target business data from the sequence business data according to the sorting order and put it into the target business object.
  • These unconverted business data can indicate the potential preferences of the target business object (the target business object has a high probability of converting the unconverted items in the unconverted business data), sort and deliver the business data containing unconverted items according to preferences , so that the delivered business data is in line with the preferences of the target business object, so it is more accurate and reduces invalid delivery, so it can save processing resources.
  • FIG. 1 is a network architecture diagram provided by an embodiment of the present application
  • Fig. 2a-Fig. 2c are schematic diagrams of a scene for determining the sorting factor provided by the embodiment of the present application;
  • FIG. 3 is a schematic flow diagram of a business data processing method provided in an embodiment of the present application.
  • Fig. 4 is a schematic flowchart of another business data processing method provided by the embodiment of the present application.
  • Fig. 5 is a schematic flow chart of recommending similar business data based on the browsing feedback data of the target business object provided by the embodiment of the present application;
  • FIG. 6 is a system structure diagram provided by an embodiment of the present application.
  • FIG. 7 is a schematic structural diagram of a business data processing device provided in an embodiment of the present application.
  • Fig. 8 is a schematic structural diagram of a computer device provided by an embodiment of the present application.
  • FIG. 1 is a network architecture diagram provided by an embodiment of the present application.
  • the network architecture may include a service server 1000 and a user terminal cluster.
  • the user terminal cluster may include one or more user terminals, and the number of user terminals is not limited here.
  • a plurality of user terminals may include a user terminal 100a, a user terminal 100b, a user terminal 100c, ..., a user terminal 100n; as shown in Figure 1, a user terminal 100a, a user terminal 100b, a user terminal 100c, ...,
  • the user terminals 100n can respectively be connected to the service server 1000 through a network, so that each user terminal can perform data interaction with the service server 1000 through the network connection.
  • each terminal device (referred to as a user terminal) as shown in FIG. 1 can be installed with a target application. 1000 perform data interaction, so that the service server 1000 can receive service data from each terminal device.
  • the target application may include an application having a function of displaying text, image, audio, video and other data information.
  • the application may be an application that supports service data delivery, for example, the application may be a multimedia application (for example, a video application), an entertainment application (for example, a game application), and a social application.
  • the service data in the embodiment of the present application may be media data (eg, advertisement data), and the following will be described by taking the service data as advertisement data as an example.
  • the service server 1000 in the embodiment of this application can create objects from the service according to these applications (the service creation object can refer to the application account of the user who creates the advertisement data in the application; the service creation object can also refer to the account corresponding to the user who creates the advertisement data terminal device; wherein, the user who creates the advertisement data can also be referred to as a service creation user or an advertiser), obtains the advertisement data created by the service creation user to be placed; the service server 1000 in the embodiment of the present application can also be based on These applications obtain the delivery logs (advertising logs, That is, the relevant delivery data of the advertisement data), each delivered advertisement data can correspond to a delivery log (or a delivery log can contain delivery related data of all the delivered advertisement data), and the delivery log can include The exposure amount of each delivered advertisement data (the object identifier of the business object that clicks and plays the advertisement data after delivering the advertisement data to the user group (that is, the application account in the application of the user whose advertisement data is delivered) number) and real-time conversion number
  • the delivered advertisement data may be referred to as delivered service data (delivered advertisement data), and the delivered service data will be referred to as delivered advertisement data for illustration below.
  • the business server 1000 After the business server 1000 obtains the advertisement log of the delivered advertisement data, it can also acquire the set of delivered business objects, the delivered advertisement data clicked and played by each business object, and the conversion behavior of each business object can also be determined (For example, collections, order purchases, and store visits, etc.) delivered advertising data.
  • the business object can be called the target business object, and the delivered business data can be called unconverted advertising data (also called unconverted advertising data). transform business data).
  • the service server 1000 can establish the unconverted browsing behavior characteristics of the target business object for the unconverted advertisement data, and determine the relationship between the target business object and the unconverted items (that is, the items included in the unconverted advertisement data) according to the unconverted browsing behavior characteristics.
  • the business data (advertising data) containing the same unconverted item can be sorted based on the sorting factor, and then the target business data (target advertising data) is selected from the sorted business data in sequence and delivered to the target business object.
  • the target business data target advertising data
  • for establishing the unconverted browsing behavior characteristics of the target business object for the unconverted advertising data, and determining the ranking factor corresponding to the target business object and the unconverted items according to the unconverted browsing behavior characteristics, and putting the target business data into the business For the specific implementation manner of transferring data to the target business object, refer to the description corresponding to the subsequent FIG. 3 .
  • the advertisement data containing the same item can be placed for the business object, because these advertisement data are in line with the preferences of the business object.
  • the business object It is likely to click and watch the delivered advertisement data and generate conversion behavior, thereby improving the conversion rate of the advertisement data.
  • one terminal device may be selected as a target terminal device among multiple terminal devices (the target terminal device may be a terminal device corresponding to a service creation user, or a terminal device corresponding to a service delivery platform), and the terminal device may Including: smartphones, tablet computers, laptop computers, desktop computers, smart TVs, smart speakers, desktop computers, smart watches, and vehicle-mounted devices that carry multimedia data processing functions (such as video data playback functions, music data playback functions) Smart terminals, but not limited to this.
  • the user terminal 100a shown in FIG. 1 may be used as the target terminal device, and the above-mentioned target application may be integrated in the target terminal device. For data interaction.
  • the service data processing method provided in the embodiment of the present application can be executed by a computer device (referred to as an electronic device), and the computer device includes but is not limited to a terminal device or a service server.
  • the business server can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers, and can also provide cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud Cloud servers for basic cloud computing services such as communications, middleware services, domain name services, security services, content distribution network (CDN, Content Delivery Network), and big data and artificial intelligence platforms.
  • terminal device and the service server may be connected directly or indirectly through wired or wireless communication, which is not limited in this embodiment of the present application.
  • the above-mentioned computer equipment can be a node in a distributed system, wherein the distributed system can be a blockchain system, and the area The block chain system can be a distributed system formed by connecting multiple nodes through network communication.
  • nodes can form a peer-to-peer (P2P, Peer To Peer) network
  • P2P protocol corresponding to the peer-to-peer network is an application layer protocol running on the Transmission Control Protocol (TCP, Transmission Control Protocol).
  • TCP Transmission Control Protocol
  • any form of computer equipment such as electronic equipment such as business servers and terminal equipment, can become a node in the blockchain system by joining the peer-to-peer network.
  • TCP Transmission Control Protocol
  • Blockchain is a new application model that uses computer technologies such as distributed data storage, point-to-point transmission, consensus mechanism, and encryption algorithm. It is used to organize data in chronological order and encrypt them into ledgers so that they cannot be tampered with or forged. , and data verification, storage and update can be performed at the same time.
  • the computer device is a block chain node
  • the data in the embodiment of the application can be authentic and Security, so that the results of related business data processing based on these data can be more reliable than the business data processing results obtained without combining the blockchain.
  • FIG. 2a-FIG. 2c are schematic diagrams of a scenario for determining a sorting factor provided by an embodiment of the present application.
  • the user terminal 100a-100e shown in FIG. 2a-FIG. 2b belong to each terminal device corresponding to FIG. 1 above.
  • the scenarios shown in Figures 2a-2c are an example of delivering an advertisement data set 20 to a business object set 200, wherein the advertisement data set 20 may include advertisement data 2001, advertisement data 2002, advertisement data 2003, ..., advertising data 2009, a total of 9 pieces of advertising data; among them, advertising data 2001, advertising data 2002, advertising data 2003, advertising data 2004 and advertising data 2008 all contain the business item "liquid foundation"; advertising data 2005 and advertising data 2006 Both include the business item "pure milk"; the advertisement data 2007 and the advertisement data 2009 include the business item "biscuit".
  • each advertisement data in the advertisement data set is already delivered advertisement data, so each advertisement data is referred to as delivered advertisement data hereinafter.
  • the business object set 200 may include business object a1, business object a2, business object a3, business object a4, and business object a5, and the clicks of each business object on these delivered advertisement data may be obtained Behavior (the click behavior means that the business object clicks to play the advertisement data) and conversion behavior (the conversion behavior means that the business object generates an order or purchase behavior after clicking to play a certain business data (that is, the order is placed or the business data is purchased) The business item promoted in )); Then, for each business object, it can be determined that the delivered advertisement data has click behavior and non-conversion behavior.
  • Behavior the click behavior means that the business object clicks to play the advertisement data
  • conversion behavior means that the business object generates an order or purchase behavior after clicking to play a certain business data (that is, the order is placed or the business data is purchased) The business item promoted in )
  • the object identifier of the business object (used to represent the business object, the object identifier can be an object name, an object identity (ID, Identity) Document), etc.), the advertisement identifier of the advertised data (used to represent the advertisement data, the advertisement identifier can be the advertisement name, the advertisement ID, etc.), whether there is a click behavior, whether there is a conversion behavior, and the recording time of the click behavior, etc.
  • the business object a1 and the delivered advertisement data 2001 are taken as examples for illustration.
  • the data group for the business object a1 and the delivered advertisement data 2001 can be [a1, 2001, 1, 0, 14 :00], where the third value 1 in the data group can be used to represent that the business object a1 has clicked on the delivered advertisement data 2001, and the fourth value 0 in the data group can be used to represent the business object a1 has no conversion behavior for the delivered advertisement data 2001
  • the time information 14:00 in this data group can be 14:00 on August 7, 2021, and this time information can be used to represent the delivered advertisement data of business object a1
  • the time when the advertisement data 2001 generates a click behavior that is, the business object a1 clicks to play the delivered advertisement data 2001 at 14:00 on August 7, 2021.
  • whether the business object a1 has a click behavior on the delivered advertisement data can be understood as the business output status of the business object a1 on the delivered advertisement data, for example, if the business object a1 has a click behavior on the delivered advertisement data 2001, it is the business object a1 clicks to play the delivered advertisement data 2001, then the business output status of the business object a1 for the delivered advertisement data 2001 is the output status; and if the business object a1 has no click behavior for the delivered advertisement data 2001, it is a business The object does not click to play the delivered advertisement data 2001, and the service output status of the business object a1 for the delivered advertisement data 2001 is not output.
  • Whether or not the business object a1 has a conversion behavior for the delivered advertisement data can be understood as the business conversion status of the business object for the delivered advertisement data. For example, if the business object a1 has a conversion behavior for the delivered advertisement data, then the business object a1 has The business conversion status of the data is the converted status; and if the business object a1 has no conversion behavior for the delivered advertising data, the business conversion status of the business object a1 for the delivered advertising data is the unconverted status.
  • the above data group [a1, 2001, 1, 0, 14:00] for the business object a1 and the delivered advertisement data 2001 may be determined as the browsing behavior characteristic of the business object a1 for the delivered advertisement data 2001 .
  • the data group of all delivered advertisement data for each business object can be constructed, that is, the browsing behavior characteristics of each business object for all delivered advertisement data can be constructed.
  • each browsing behavior characteristic includes the object identifier of each business object, the advertisement identifier of the delivered advertisement data, whether there is a click behavior, whether there is a conversion behavior .
  • the recording time of the click behavior (if no click behavior occurs, the recording time can be a specified value, such as a null value), then through the characteristics of each browsing behavior, it can be determined for each business object that there is a click behavior at the same time Served ad data with non-converting behavior.
  • the delivered advertisement data with click behavior but no conversion behavior for the business object may be referred to as unconverted business data for the business object (will be referred to as unconverted advertisement data in FIG. 2a-FIG. 2c).
  • the business object a1 is taken as an example to illustrate the unconverted advertisement data corresponding to the business object.
  • each business object can be used as a target business object, and the browsing behavior characteristics of each target business object with respect to the respective unconverted advertisement data can be determined.
  • browsing behavior characteristics of a target business object for a piece of unconverted advertisement data may be referred to as unconverted browsing behavior characteristics.
  • business object a3 and business object a4 as well as delivered advertisement data set 20b, delivered advertisement data set 20c, and delivered advertisement data set 20d, etc. are shown in the form of ellipsis defined.
  • the target business object as the business object a1 (hereinafter referred to as the target business object a1) as an example (because the delivered advertisement data set 20a is the unconverted advertisement data corresponding to the business object a1, then the following will be
  • Each delivered advertisement data in the delivered advertisement data set 20a is referred to as unconverted advertisement data) indicating the unconverted browsing behavior characteristics of the target business object with respect to the unconverted advertisement data.
  • the browsing behavior of the target business object a1 with respect to the unconverted advertisement data 2001 is characterized by the data group [a1, 2001, 1, 0, 14:00], and the data group [a1, 2001, 1, 0, 14:00] can be used as The target business object a1 is directed at the unconverted browsing behavior characteristics 2000a of the unconverted advertisement data 2001; similarly, the target business object a1 is data group [a1, 2002, 1, 0, 14: 10], the browsing behavior characteristic of the target business object a1 for the unconverted advertisement data 2003 is the data group [a1, 2003, 1, 0, 14:20], the browsing behavior characteristic of the target business object a1 for the unconverted advertisement data 2004 is the data Group [a1, 2004, 1, 0, 14:30], target business object a1
  • the browsing behavior of unconverted advertisement data 2005 is characterized by data group [a1, 2005, 1, 0, 14:40], target business object a1
  • the browsing behavior for the unconverted advertisement data 2006 is characterized by the data group [a1, 2006, 1, 0, 14:50] (the time information
  • advertisement identification can be performed on the unconverted advertisement data corresponding to the target business object a1 to determine the business items included in the unconverted advertisement data (because the advertisement data is an unconverted advertisement, then for Understand that the business items included in the unconverted advertising data will be referred to as unconverted items in the following), and then the advertisement identifiers of the unconverted advertising data contained in each unconverted browsing behavior feature will be replaced by the business items included in the unconverted advertising data
  • the item identification of the target business object a1 can obtain the target browsing behavior characteristics of each business item.
  • the unconverted advertisement data is taken as an example to illustrate the target browsing behavior characteristics.
  • the advertisement data 2001, advertisement data 2002, advertisement data 2003, advertisement data 2004 and advertisement data 2008 all contain the business item "liquid foundation", advertisement data 2005, advertisement data 2006, advertisement data Both 2007 and advertisement data 2008 contain the business item "pure milk"; thus, through advertisement identification, it can be determined that the unconverted advertisement data 2001 contains the untransformed item "foundation".
  • the item identifier of "foundation” is FDY001
  • the data set [a1, 2001, 1, 0, 14:00] can be The advertisement ID 2001 in is replaced by the item ID FDY001 of "Foundation Foundation", so that the target browsing behavior feature 2000a' of the target business object a1 for the unconverted item "Foundation Foundation” can be obtained as [a1, FDY001, 1, 0, 14 :00].
  • the target browsing behavior feature 2000b' of the target business object a1 for the unconverted item "liquid foundation” can be determined (that is, the data group [a1, FDY001, 1, 0, 14:10], corresponding to the unconverted browsing behavior feature 2000b), the target browsing behavior feature 2000c' for the unconverted item "Foundation Foundation” (that is, [a1, FDY001, 1, 0, 14:20], corresponding to the unconverted browsing behavior feature 2000c), and the target browsing behavior feature 2000c for the unconverted item "Foundation Foundation”
  • the target browsing behavior feature 2000d' of "liquid” namely [a1, FDY001, 1, 0, 14:30], corresponding to the unconverted browsing behavior feature 2000d
  • the target browsing behavior feature 2000e' for the unconverted item "pure milk” ie [a1, CN001, 1, 0, 14:40], corresponding to the unconverted browsing behavior feature 2000e, CN001 is the item identification of "pure milk”
  • the target browsing behavior characteristics belonging to the same business item can be obtained from the target browsing behavior characteristics of the target business object a1 for each business item (ie, the target browsing behavior features with the same item identifier).
  • the ranking factor corresponding to the target business object a1 and the business item can be determined (the ranking factor can be used to represent the degree of interest of the target business object for a certain business item; that is In other words, the ranking factor corresponding to a target business object and a certain business item can be understood as the interest degree of the target business object to this business item.
  • the ranking factor of a target business object and a certain business item is related to the target
  • the number of clicks of the business object on the business item is associated, and the number of clicks can indicate the degree of interest of the target business object in the business item. If the target business object has more clicks on a certain business item, it can indicate that the target business object The higher the degree of interest in the business item, the greater the ranking factor corresponding to the target business object and the business item).
  • the process of determining the ranking factor will be illustrated below with an example.
  • take the untransformed item "Foundation Foundation” as an example to illustrate the process of determining the ranking factor; 14:10], data group [a1, 2003, 1, 0, 14:20] and data group [a1, 2004, 1, 0, 14:30] are target business object a1 for the unconverted item "liquid foundation” browsing behavior characteristics, so that the total number of clicks of the target business object a1 on the unconverted item "liquid foundation” can be determined according to these browsing behavior characteristics (a browsing behavior characteristic corresponds to a click behavior, then the total number of clicks can be The total number of browsing behavior features is 4); then, the earliest record time (that is, the earliest time when the target business object a1 clicks on the unconverted item "liquid foundation”) can be obtained from these browsing behavior features , can also be called the earliest click time), and the earliest record time is 14:00 (that is, 14:00 on August 7, 2021); at the same time, the current time can also
  • the ranking factor 1 corresponding to the target business object a1 and the unconverted item "foundation” is associated with the number of clicks of the unconverted item "foundation” by the target business object a1.
  • the earliest record time and the current time please refer to the subsequent descriptions corresponding to FIGS. 3-4 .
  • the ranking factor 2 corresponding to the target business object a1 and the unconverted item "pure milk” can also be determined.
  • the advertised data that is, the advertisement data 2001, the advertisement data 2002
  • the advertisement data to be placed including the untransformed item “liquid foundation” and the unconverted item “pure milk” can also be obtained.
  • the set of candidate advertisement data may refer to a set of advertisement data waiting to be delivered.
  • the advertised data (that is, the advertised data to be sorted) are all determined to include the untransformed item "liquid foundation” and the unconverted item “pure milk”.
  • the embodiment of the present application can refer to these advertisement data as unbroadcast advertisement data, unclicked advertisement data or unoutput advertisement data), if the unbroadcast advertisement data is advertisement data 2009, then according to the sorting Factor 1 and sorting factor 2 can sort the advertisement data containing the untransformed item "liquid foundation" and the unconverted item "pure milk” and the unexported advertisement data 2009 to obtain the serial advertisement data, and follow the The sorted result selects the target advertisement data and delivers it to the target business object a1.
  • the sorting manner and the manner of selecting the target advertisement data refer to the subsequent description corresponding to FIG. 3 .
  • the embodiment of the present application can determine the unconverted items based on the unconverted browsing behavior characteristics The sorting factor of the item, and then sort the advertisement data containing the unconverted item and the unclicked advertisement data based on the sorting factor.
  • the increase will include the unconverted
  • the advertising data of the target business object a1 is put into the target business object a1, because the target business object a1 is interested in, then compared with the probability of the target business object a1 generating conversion behavior on other advertising data that does not contain unconverted items, the target business object a1 has Advertising data with unconverted items has a higher probability of generating conversion behavior, which can increase the conversion rate.
  • accurate delivery is performed according to the preference of each business object, so the conversion rate can be improved.
  • advertisement identifier such as 2001
  • object identifier such as a1
  • item identifier such as FDY001
  • conversion The value of the behavior (1 or 0)
  • recording time stamp such as 14:00
  • browsing behavior characteristics such as the data group [a1, 2002, 1, 0, 14:10]
  • Fig. 3 is a schematic flowchart of a business data processing method provided by an embodiment of the present application, wherein the business data processing method can be executed by electronic devices such as computer equipment, and here the business data processing method is executed by computer equipment case is described as an example.
  • the computer device here may refer to a service server (such as the service server in the above-mentioned FIG. 1 ), or may refer to a terminal device (such as any terminal device in the terminal device cluster in the above-mentioned FIG. 1 ).
  • the flow of the business data processing method may at least include step S101-step S103, and each step will be described separately below.
  • service data may refer to media data (such as advertisement data), delivery may refer to recommended processing such as exposure, and delivered service data may refer to media data that has been delivered to business objects.
  • the business object may refer to the bound account in the target application of the business user who uses the terminal device to run the target application (such as entertainment application, social application, video application, etc.).
  • the business user can use the bound account to log in to the target application, and the target
  • the application can also determine whether the business user is logged in by binding the account, obtain relevant behavior data of the business user in the target application, and so on.
  • the above-mentioned target applications such as entertainment applications, social networking applications, and video applications can be used as advertising delivery platforms, and advertisers who create advertising data (also called business creation objects) can use Advertising data is delivered to the advertising delivery platform.
  • advertising data business user groups can be targeted for delivery. Different types of advertising data can be targeted to different business user groups in the target application. That is to say, Different types of advertisement data will be delivered to different sets of business objects in the target application (called bound accounts corresponding to business user groups).
  • advertisement data can be called the delivered advertisement data
  • each business user in this part of the business user group can play and watch the delivered advertisement data
  • each business After playing and watching users can also consume (such as purchase the products in the advertised data), download and other behaviors, and the behavior of business users to consume or download the advertised data can be understood as conversion behavior. Consumption or download of a delivered advertisement data can be understood as a conversion.
  • business user group A which can be called business user set A
  • business user set A For a certain advertising data, if there are 50 users in the business user set A If a certain product is purchased through the advertisement data, the number of conversions of the business user set A for the advertisement data is 50.
  • the computer equipment can obtain the relevant data of the advertising data (including the playback and viewing behavior of the business user, conversion behavior, viewing duration, name of the advertising data) through the bound account of the business user group (that is, the business object). , the type of advertising data and the exposure of advertising data, etc.).
  • the service output status in the embodiment of the present application can be understood as whether the business user has the behavior of playing and watching the advertisement data that has been placed (playing and watching it means It can be understood that the delivered advertisement data is output to the business user), if a certain business user plays and watches a certain delivered advertisement data, then the business object corresponding to the business user will output the business status of the delivered advertisement data That is, the output state (that is to say, the business user corresponding to the business object has watched the advertisement data); If the advertising data is delivered, the business output status of the business object corresponding to the business user for the delivered advertising data can be the non-output status.
  • a computer device can determine whether a business user has a conversion behavior for a certain advertised data through the bound account of a business user group (also called a business user set), and the business conversion status in this embodiment of the application is It can be understood as whether business users have conversion behaviors on the delivered advertising data. If a business user has a conversion behavior for a certain delivered advertisement data, the business conversion status of the business object corresponding to the business user for the delivered advertisement data can be the converted state; If there is no conversion behavior in the delivered advertisement data, the business conversion status of the business object corresponding to the delivered advertisement data can be an unconverted status.
  • a business user's behavior of playing and viewing certain delivered advertisement data may be generated by a trigger operation performed by the business user on the display interface of the delivered advertisement data in the terminal device; wherein, the trigger operation It may include contact operations such as click or long press, and may also include non-contact operations such as voice or gesture, which will not be limited here.
  • the trigger operation It may include contact operations such as click or long press, and may also include non-contact operations such as voice or gesture, which will not be limited here.
  • the computer device can record the data of the delivered advertisement.
  • the computer device When the click operation is generated, when the business user clicks, the computer device will play the delivered business data to the business object corresponding to the business user, then the business output status of the business object for the delivered advertisement data will also be determined by The non-output state is changed to the output state, so the click operation can also be understood as a state change operation, and the recording time can also be referred to as the record time stamp for the state change operation.
  • the data related to the delivered advertisement data acquired by the above computer equipment may be referred to as a delivery log of the delivered advertisement data (also called an advertisement log), and the delivery log may also include the aforementioned record time stamp.
  • the delivery log may also include the record time stamps of the service object sets respectively for the N delivered service data.
  • the set of business objects includes the business object M i (i can be any value used to characterize the subscript, which is an index variable, for example, i can be a positive integer or a fraction, etc.), and N delivered business data include Putting business data G b (b can be any value used to characterize the subscript, it is an index variable, for example, b can be a positive integer or a fraction, etc.)
  • the record time stamp of b refers to the operation record time stamp of the business object M i ’s state change operation on the delivered business data G b , where the state change operation can be used to indicate that the business object M i ’s business on the delivered business data G b
  • the output state is changed from the non-output state to the output state.
  • the computer equipment aims at the business output status and business conversion status of the N delivered business data respectively, and the process of establishing the unconverted browsing behavior characteristics of the target business object for the unconverted business data includes: the computer equipment can obtain the business object M The object identifier of i and the service identifier of the delivered business data G b ; the computer device can use the object identifier of the business object M i , the service identifier of the delivered business data G b , and the business object Mi for the delivered business data G b
  • the output status, the business conversion status of the business object M i for the delivered business data G b , and the data group composed of the business object M i ’s record time stamp for the delivered business data G b are determined as the business object M i for the delivered business Browsing behavior characteristics of the data Gb ; then, the computer device can determine the unconverted browsing behavior characteristics of the target business object for unconverted business data from the browsing behavior characteristics of the business object set for N delivered business data.
  • the computer device determines the unconverted browsing behavior characteristics of the target business object for the unconverted business data in the browsing behavior characteristics of the business object set for the N delivered business data, including: the computer device can use the business object set for the N Among the browsing behavior characteristics of the delivered business data, the browsing behavior characteristics whose business output status is output and whose business conversion status is unconverted are determined as unconverted browsing behavior characteristics; The business object corresponding to the included object identifier is determined as the target business object, and the delivered business data corresponding to the business identifier included in the unconverted browsing behavior characteristic is determined as unconverted business data.
  • the business output status for example, whether there is a playback and viewing behavior
  • business conversion status whether there is a conversion behavior
  • the object identifier of the business object for characterizing Business object, the object identifier can be object name, object ID, etc.
  • the advertisement identifier of the delivered business data used to represent the business data, the business identifier can be business name, business ID, etc.
  • business output status business conversion status and Generate data such as time stamps of playback and viewing behaviors
  • the data group can be the browsing behavior characteristics of the service object for the delivered service data.
  • the data groups whose business output state is the output state and the business conversion state is the unconverted state can be extracted, and these extracted data groups can be called unconverted browsing behavior characteristics (or is called the unconverted data group), the business object contained in the unconverted browsing behavior feature can be called the target business object, and the delivered business data contained in the unconverted browsing behavior feature can be called the unconverted business data .
  • the business object A played and watched the delivered business data A, but did not generate conversion behavior for the delivered business data A
  • the business object A can be called the target business object
  • the delivered business data A can be called
  • the data group composed of the business object A and the delivered business data A may be called the unconverted browsing behavior feature.
  • the method of determining the target business object, non-converted business data and non-converted browsing behavior characteristics will be illustrated below with examples. Take the business object set including business object A, and the delivered business data including delivered advertisement data A as an example; At 19:06, the delivered advertisement data A was played and watched, and the business output status of the business object A for the delivered advertisement data A was output. After playing and watching at 19:06 on August 7, 2021, there was no conversion behavior.
  • the value 1 is used to represent the output state and the converted state
  • the value 0 is used to represent the non-output state and the non-transformed state
  • the object identifier of the business object A namely A
  • the advertisement identifier of the delivered advertisement data A namely A
  • business conversion status (1 or 0)
  • record time stamp (19:00, 19:06) can form 2 data groups, including data group [A, A, 1, 0 , 19:00], data group [A, A, 1, 0, 19:06].
  • the computer device when determining the unconverted browsing behavior characteristics, because the data group [A, A, 1, 0, 19:00] and the data group A, A, 1, 0, 19:06]
  • the business output status is the output status
  • the business conversion status is the untransformed status at the same time, so that the computer device can determine the two data groups as the untransformed browsing behavior characteristics, and the business object a can be determined as the target business object, and the The served advertisement data A is determined to be unconverted business data.
  • unconverted browsing behavior characteristics determine the sorting factor corresponding to the target business object and the unconverted item, sort the business data containing the unconverted item according to the sorting factor, obtain the sequence business data, and start from the sequence business according to the sorting order Select the target business data in the data, and deliver the target business data to the target business object; unconverted business data includes unconverted items.
  • the service data may be advertisement data, and each advertisement data is used to promote a certain product, and the product promoted by each advertisement data may be called a service item.
  • Each business data in the embodiment of this application can contain one (optional, or multiple, when the business data contains multiple business items, one business item can be selected as the business item contained in the business data) Items) business items, and the business items included in the untransformed business data can be called unconverted items.
  • the sorting factor corresponding to the target business object and the unconverted item can be determined.
  • the ranking factor can be used to represent the degree of interest of the target business object in a certain business item (such as an unconverted item).
  • an unconverted browsing behavior feature actually includes an object identifier of a target business object and a business identifier of unconverted business data, while one unconverted business data contains an unconverted item, and multiple unconverted business data contains may all contain the same unconverted item; and in the embodiment of the present application, the computer device can determine the The ranking factor corresponding to the target business object and the unconverted item.
  • the relationship between the target business object a1 and the unconverted browsing behavior can be determined according to the unconverted browsing behavior characteristics 2000a, The sorting factor 1 corresponding to the conversion item "Foundation Foundation" is common.
  • the computer device can determine the ranking factor corresponding to a target business object and an unconverted item.
  • the ranking factor refer to the subsequent description corresponding to FIG. 4 .
  • the computer device can sort the business data containing the unconverted items based on the sorting factor corresponding to the target business object and a certain unconverted item, select the target business data according to the sort order, and put it into the target business object.
  • the business data containing the untransformed items includes Q target unconverted business data, the Q target unconverted business data all contain the same unconverted item P x , and the Q target unconverted business data Including the target unconverted business data R s (representing each target unconverted business data in the Q target unconverted business data, s is any value used to characterize the subscript, such as s is a positive integer whose value range is 1 to Q ) as an example for illustration.
  • the computer equipment sorts the business data containing unconverted items, and the process of obtaining the sequence business data includes: the computer equipment can obtain the total delivery amount of the target untransformed business data R s for the set of business objects; according to the total delivery amount, and the business object Collecting the real-time conversion quantity for the target unconverted business data R s , the real-time delivery conversion rate corresponding to the target unconverted business data R s can be determined; then, the computer equipment according to the ranking factor corresponding to the target business object and the unconverted item P x , As well as the real-time delivery conversion rates corresponding to the Q target unconverted business data respectively, the business data containing unconverted items P x can be sorted to obtain sequence business data.
  • the total amount of delivery can actually be understood as the total exposure (it can also be understood as the number of exposures of delivered business data, for example, if a certain business data is exposed to one thousand business objects, the business data can be called delivered Business data, the exposure of the delivered business data is 1,000 times);
  • the real-time conversion number can be understood as the actual (real) number of business objects that undergo conversion behavior after the business data is delivered, the real-time conversion number It is also an indicator to measure the effect of business data delivery. It refers to the number of business users who click on business data (such as advertising data) and become a valid activation, registration or paying user in a set of business users.
  • the real-time conversion ratio can also be obtained through the number of real-time conversions.
  • the real-time conversion rate that is, the actual number of conversions (quantity) of the advertising data divided by the number of clicks (number of clicks) of the advertising data).
  • the real-time delivery conversion rate of the delivered business data can be obtained (for example, the real-time delivery conversion rate can be obtained by dividing the real-time conversion quantity by the total delivery volume, It can also be called real-time exposure conversion rate).
  • the computer equipment when the business data containing unconverted items also includes the business data to be delivered including unconverted items P x , the computer equipment is based on the ranking factor corresponding to the target business object and unconverted items P x , and The real-time delivery conversion rates corresponding to the Q target untransformed business data respectively, sort the business data containing the unconverted items P x , and the process of obtaining the sequence business data includes: the computer equipment can obtain the pending items containing the unconverted items P x The business data to be delivered, and the estimated delivery conversion rate corresponding to the business data to be delivered; and then, the computer equipment is calculated according to the real-time delivery conversion rate corresponding to the delivered business data to be sorted, and the estimated delivery conversion rate corresponding to the business data to be delivered In the order of size, the business data containing unconverted items P x can be sorted, so that the initial sequence business data can be obtained; in the initial sequence business data, the K with the largest estimated delivery conversion rate can be obtained in order (K can
  • the computer device can obtain the delivered service data to be sorted including the unconverted item P x from the N delivered service data; among them, the delivered service data including the unconverted item P x (that is, to be The sorted delivered business data) includes Q target unconverted business data; in addition, the computer device can determine the sorted delivered business data and the pending delivered business data as business data containing unconverted items P x .
  • K business data including business data R t (t is a positive integer) as an example to illustrate the acquisition process of sequence business data
  • K business data including business data R t means that business data R t is the K Any of the business data.
  • the computer equipment sorts the K pieces of business data according to the sorting factor corresponding to the target business object and the unconverted item P x , and the expected conversion unit input resources, estimated output rate, and estimated conversion rate corresponding to the K pieces of business data
  • the process of obtaining the sequence business data includes: the computer device releases resources according to the expected conversion unit corresponding to the business data R t , the estimated output rate and the estimated conversion rate, and the computer device can determine the resource consumption of the initial input unit corresponding to the business data R t ;
  • the ranking factor corresponding to the target business object and the untransformed item P x , and the initial delivery unit consumption resource corresponding to the business data R t are calculated and processed, and the real-time delivery unit consumption resource corresponding to the business data Rt is obtained; then, the
  • the computer device can obtain from the N pieces of delivered business data that the target business object has not generated the behavior of playing and viewing (that is, no click behavior has been generated, and the target business object is aimed at the business of the delivered business data.
  • the delivered business data whose output status is not output state) can be used as the unoutputted business data (also called unbroadcasted business data, or unclicked business data); That is to say, the business data including unconverted items also includes unexported business data.
  • the computer device can sort the business data containing the untransformed item P x together with the unoutput business data, that is, the N delivered business data can also include the unoutput business data, then the computer device can sort the untransformed business data according to the K Real-time delivery unit consumes resources in order of size, sorts K business data, and the process of obtaining sequence business data includes: computer equipment obtains the expected conversion unit delivery resources corresponding to unoutput business data, estimated output rate, and estimated Conversion rate; according to the expected conversion unit corresponding to the non-output business data, the resource consumption of the initial delivery unit corresponding to the unoutput business data, the estimated output rate, and the estimated conversion rate are determined; according to the initial delivery unit corresponding to the unoutput business data Consume resources, and the size order between K real-time delivery units consume resources, sort K business data and unoutput business data, and obtain sequence business data.
  • the computer device can obtain the delivered advertisement data (including the delivered advertisement data 2001, the delivered advertisement data 2001, and the Advertisement Data 2002, Advertisement Data 2003, Advertisement Data 2004 and Advertisement Data 2008), including Advertisement Data of the unconverted item "Pure Milk” (including Advertisement Data 2005 and Advertisement Data 2006 ). Then the obtained delivered advertisement data can be called delivered service data including unconverted items P x (that is, to-be-sorted delivered service data including unconverted items P x ).
  • the computer device can also be included in the candidate business data set (that is, the set composed of candidate business data waiting to be delivered, none of which has been placed in the business object set, and the candidate business data can also be It is called the business data to be delivered, and the candidate business data set is the candidate advertisement data set.
  • the candidate business data set may contain historical business data that has not been delivered, or newly created ones that have not been sorted and have not been delivered.
  • New business data, historical business data and new business data can be referred to as candidate business data), obtain the business data to be launched that includes the unconverted item P x (for example, include the unconverted item "liquid foundation” or unconverted
  • the advertisement data to be delivered of the item "pure milk” can be composed of the advertisement data to be delivered including the unconverted item "liquid foundation” and the above-mentioned business data to be sorted and delivered including the unconverted item "liquid foundation” to form an unconverted
  • the business data of the item "Foundation Foundation"; the advertising data to be delivered containing the unconverted item "Pure Milk” and the business data to be sorted and delivered containing the unconverted item “Pure Milk” can be combined to form the unconverted Milk” business data).
  • the computer device can obtain the delivered advertisement data 2001, delivered advertisement data 2002, delivered advertisement data 2003, delivered advertisement data 2004, delivered advertisement data 2008, delivered advertisement data 2005 and delivered The corresponding real-time delivery conversion rates in the advertisement data 2006; at the same time, because the pending advertisement data including the unconverted item "liquid foundation" and the unconverted item "pure milk" have not been delivered, there is no real-time delivery conversion rate.
  • the computer device can obtain the estimated output rate of each of the advertising data to be delivered (also called the predicted click through rate (Predict Click Through Rate, PCTR), the estimated click through rate refers to the advertising data in the
  • the online advertising system estimates the probability that the advertisement data will be clicked; that is, after the computer device puts an advertisement data into a certain business object set, the estimated business object set will be The ratio between the number of business objects that will click on the advertisement data and the total number of business objects in the business object collection) and the estimated conversion rate (Predict Conversion Rate, PCVR; the estimated conversion rate refers to the advertising data being clicked in a certain situation
  • the online advertising system estimates the probability that the clicked advertisement data will be converted; that is, after a computer device puts an advertisement data into a certain set of business objects, the advertisement data is clicked on some business objects.
  • the ratio of the number of objects that will convert the advertisement data to the total number of clicked business objects) can be based on the estimated click-through rate and the estimated click-through rate corresponding to each advertisement data to be delivered.
  • the estimated conversion rate is determined to determine the estimated delivery conversion rate corresponding to the advertisement data to be delivered (for example, the estimated click rate is multiplied by the estimated conversion rate to obtain the estimated delivery conversion rate).
  • the computer equipment can sort the business data including the unconverted item "foundation liquid” and the unconverted item “pure milk” according to the order of the real-time delivery conversion rate and the estimated delivery conversion rate , if the pending business data containing the unconverted item "liquid foundation" or the unconverted item “pure milk” is the pending business data 20010, sort it in descending order of real-time delivery conversion rate and estimated delivery conversion rate
  • the obtained initial sequence business data is ⁇ 2008 delivered advertisement data, 2001 delivered advertisement data, 2006 delivered advertisement data, 2003 delivered advertisement data, 20010 delivered advertisement data, 2002 delivered advertisement data, 2002 delivered advertisement data, 2002 delivered advertisement data Data 2004, advertised data 2005 ⁇ ;
  • the computer device can obtain the first K placed business data in the initial sequence of business data according to the needs of the business scenario, taking K as 5 as an example,
  • the first 5 delivered advertisement data can be extracted: ⁇ 2008 delivered advertisement data, 2001 delivered advertisement data, 2006 delivered advertisement data, 2003 delivered advertisement data, 20010 delivered advertisement data ⁇ .
  • the computer device can obtain the expected conversion unit of each business data in the above K business data (it can refer to the price that advertisers bid for advertising data, for example, it refers to the advertiser’s corresponding price for a conversion Expected cost price, that is, the cost of a conversion determined by the advertiser), estimated output rate and estimated conversion rate, and according to the expected conversion unit investment resources, estimated click rate, and estimated conversion rate, it can be determined
  • the initial delivery unit corresponding to each business data consumes resources.
  • the resource consumed by the initial delivery unit may refer to the cost that the advertiser needs to pay after a certain advertisement data is displayed to a thousand business users, and the resource consumed by the initial delivery unit may also be referred to as cost per thousand impressions ( Cost Per Mille, CPM).
  • CPM Cost Per Mille
  • the actual CPM may be jointly determined by the advertiser's expected conversion unit investment resources, estimated click rate, and estimated conversion rate.
  • formula (1) is an exemplary way to determine the resources consumed by the initial delivery unit.
  • CPM can refer to the actual CPM of a certain advertisement data
  • bid can refer to the expected conversion unit of the advertisement data by the advertiser (business creation user)
  • PCTR can refer to the estimated click rate of the advertisement data
  • PCVR It may refer to the estimated conversion rate of the advertisement data.
  • the estimated click-through rate of a certain advertisement data R is 0.1
  • the estimated conversion rate is 0.1
  • the expected conversion unit investment resource is 2
  • the CPM of the advertisement data can be 20 (2 ⁇ 0.1 ⁇ 0.1 ⁇ 1000);
  • advertisers need to be charged 20 yuan for every thousand impressions.
  • the real-time CPM can be used as a basis for sorting advertisement data.
  • the impact of the sorting factor needs to be added on the basis of the resources consumed by the initial delivery unit. For example, if in the scenario as shown in Fig. 2a-Fig.
  • the unexported business data of the target business object a1 that has not produced the behavior of playing and viewing is the delivered advertisement data 2009
  • the computer device can Using the above formula (1), the resource consumption of the initial delivery unit of the delivered advertisement data 2009 is obtained; then, the computer equipment can be used in the above K business data ⁇ 2008 delivered advertisement data, 2001 delivered advertisement data, 2006 delivered advertisement data, In the delivered advertisement data 2003 and the pending advertisement data 20010 ⁇ , the resources consumed by the initial delivery unit corresponding to each business data are determined.
  • the computer can obtain the ranking factor 1 corresponding to the target business object a1 and the unconverted item "liquid foundation", and the ranking factor 2 corresponding to the target business object a1 and the unconverted item “pure milk”; thus, the computer
  • the device can obtain the advertising data containing the unconverted item "Foundation Foundation” from the K pieces of business data, and calculate the resource consumption of the initial delivery unit of each advertisement data containing the unconverted item "Foundation Foundation” and the sorting factor 1 ( For example, by summing), the corresponding real-time delivery unit consumption resource (that is, the result of adding the initial delivery unit consumption resource and the sorting factor) can be obtained;
  • the resource consumption of the initial delivery unit corresponding to the advertisement data of the item "pure milk”, and the resource consumption of the real-time delivery unit corresponding to the advertisement data containing the unconverted item “pure milk” can also be determined.
  • Formula (2) is the way to determine the resource consumption of real-time delivery
  • CPM new can be used to represent the resource consumption of a real-time delivery unit of a certain advertisement data
  • bid ⁇ PCTR ⁇ PCVR ⁇ 1000 can be used to represent the resource consumption of the initial delivery unit of the advertisement data
  • quality i can be used to represent the business contained in the advertisement data The ranking factor corresponding to the item and a target business object.
  • the computer device can combine the K business data with the delivered advertisement data 2009 (or (referred to as non-output advertisement data 2009) are sorted to obtain sequence service data (for example, sequence advertisement data).
  • sequence service data for example, sequence advertisement data
  • target business data for example, target advertisement data
  • target business data can be sequentially selected from the sequenced business data according to the requirements of the business scenario and delivered to the target business object a1.
  • the computer equipment combines the above K business data ⁇ 2008, 2001, 2006, 2003, 20010 ⁇ and 2009 of the advertisement data, according to K
  • the obtained sequence business data is ⁇ the delivered advertisement data 2008, the delivered advertisement data 2001, Advertisement data 2006, advertisement data 2009, advertisement data 2003, advertisement data to be delivered 20010 ⁇
  • you can obtain the top 3 advertisement data according to the needs of the business scenario namely ⁇ advertisement delivered Data 2008, Advertisement Data 2001, Advertisement Data 2006 ⁇
  • these three advertisement data ⁇ Advertisement Data 2008, Advertisement Data 2001, Advertisement Data 2006 ⁇ can be used as target advertisement data (target business data), these three target advertisement data can be delivered to the target business object a1 in sequence.
  • the advertising data to be delivered is sorted according to the estimated click-through rate and estimated conversion rate, and then sorted according to the needs of the business scenario (estimated click-through rate and estimated conversion rate) Higher conversion rate) advertising data for delivery.
  • FIG. 4 is a schematic flowchart of another business data processing method provided by an embodiment of the present application. This process may correspond to the process of determining the ranking factor corresponding to the target business object and the unconverted item in the above-mentioned FIG. 3 . As shown in FIG. 4, the process may at least include step 401-step S403, and each step will be described below.
  • Step 401 Carry out business identification on unconverted business data to obtain a mapping relationship between unconverted business data and unconverted items; the mapping relationship is used to indicate that unconverted business data includes unconverted items.
  • the service data may be advertisement data
  • the business identification of the unconverted service data by the computer device is the advertisement identification of the advertisement data.
  • performing advertisement recognition on the advertisement data may include performing semantic content recognition on the advertisement data, and may also include performing image content recognition on the advertisement data.
  • Advertisement recognition can be carried out through speech content recognition model for semantic content recognition or image content recognition model for image content recognition, or both speech content and image content can be recognized, and then the model with higher accuracy can be selected for advertisement recognition;
  • the semantic content recognition model in the embodiment of the application can be any model that includes the function of recognizing the semantic content of advertisements
  • the model of image content recognition can be any model that includes the function of recognizing the content of advertisement image content, which is not limited in this embodiment of the application.
  • Computer equipment can obtain the mapping relationship between the delivered advertisement data and business items through semantic content recognition: R i ——> C i , where R i can be used to represent the advertisement identifier of a certain delivered advertisement data, and C i can be used In order to represent an item identifier of a certain business item, the mapping relationship may indicate that the delivered advertisement data includes the business item C i . It should be understood that the service items included in the untransformed service data may be referred to as unconverted items.
  • Step 402 obtain the item ID of the unconverted item, replace the business ID of the unconverted business data included in the unconverted browsing behavior feature with the item ID of the unconverted item, and determine the replaced unconverted browsing behavior feature as the target Business objects target browsing behavior characteristics for unconverted items.
  • the computer device can replace the service identifier of the unconverted business data in the unconverted browsing behavior feature with the item identifier of the unconverted item, thereby obtaining the target browsing behavior feature of the target business object for the unconverted item.
  • the features of unconverted browsing behavior are [a1, 2001, 1, 0, 14:00] and [a1, 2002, 1, 0, 14:10]
  • the advertisement data 2001 and the advertisement data 2002 in the unconverted browsing behavior feature both contain the unconverted item "Foundation Foundation”
  • the item identifier of the unconverted item "Foundation Foundation” is FDY001.
  • Object a1's target browsing behavior for the unconverted item "Foundation Foundation” is characterized by [a1, FDY001, 1, 0, 14:00] and [a1, FDY001, 1, 0, 14:10].
  • Step 403 according to the characteristics of the target browsing behavior, determine the ranking factor corresponding to the target business object and the unconverted item.
  • the computer device can determine the ranking factor corresponding to the target business object and the unconverted items according to the target browsing behavior characteristics of the target business object for the unconverted items.
  • the untransformed business data containing untransformed items includes Q (Q can be an integer less than or equal to N) target unconverted business data; the Q target unconverted business data all contain the same unconverted item P x (x is Any value used to characterize the subscript, x can be an integer, fraction, letter, etc.); the target browsing behavior characteristics of the target business object for unconverted items include Q target browsing behavior characteristics of the target business object for unconverted items P x , sorted
  • the factor acquisition process includes: the computer equipment can count the total number of features of Q target browsing behavior features, and the total number of features can be determined as the total output times of the target business object for unconverted items P x ; from the Q target browsing behavior features respectively Among the included record timestamps, obtain the earliest record timestamp; according to the current timestamp,
  • the unconverted advertisement data 2005 and the unconverted advertisement data 2006 both contain the same unconverted item "pure milk", then the above Q target unconverted businesses
  • the data can be the unconverted advertisement data 2005 and the unconverted advertisement data 2006;
  • the above unconverted item P x can be the unconverted item "pure milk”;
  • the above target business object's Q target browsing behaviors for the unconverted item P x The feature is the target browsing behavior feature 2000e' of the target business object a1 for the unconverted item "pure milk” (obtained through the unconverted browsing behavior feature 2000e), and the target browsing behavior of the target business object a1 for the unconverted item "pure milk” Feature 2000f' (obtained through the unconverted browsing behavior feature 2000f).
  • the computer device can determine the ranking factor corresponding to the target business object and the unconverted item according to the target browsing behavior characteristics of the target business object for the unconverted item.
  • the computer equipment is directed to the target browsing behavior characteristics (including target browsing behavior characteristics 2000a', target browsing behavior characteristics 2000b', and target browsing behavior characteristics 2000c') of the unconverted article "liquid foundation" according to the target business object a1 and the target browsing behavior feature 2000d'), the sorting factor 1 corresponding to the target business object a1 and the unconverted item "foundation" can be determined, including: the target of the computer device for the unconverted item "foundation" according to the target business object a1 Browsing behavior characteristics, determine the total number of clicks (referred to as the total output times) of the target business object a1 for the unconverted item; wherein, because a target browsing behavior characteristic corresponds to a click behavior (that is, the target business object for a certain Unconverted ad data
  • each target browsing behavior feature of the target business object a1 for the untransformed item "foundation liquid" contains the record time stamp of the business output state (that is, the time or moment when the behavior of playing and watching is generated)
  • the computer device can In the target browsing behavior feature of the target business object a1 for the unconverted item "liquid foundation", the minimum record timestamp (also called the earliest record timestamp or the earliest moment) is obtained, and the minimum record timestamp is the earliest click time (or is called the earliest click time), that is, the earliest time to play and watch); then, the computer device according to the current time (the current moment, called the current time stamp), the minimum record time stamp, and the total number of clicks, namely The sorting factor corresponding to the target business object a1 and the unconverted item "foundation" can be determined.
  • Formula (3) is the way to determine the sorting factor according to the current time, the minimum recorded timestamp and the total number of clicks, as shown below.
  • ⁇ click i in the formula (3) can be used to represent the total number of clicks of a certain target business object for a certain business item (such as the total number of clicks of the above-mentioned target business object for the unconverted item "liquid foundation"); the formula ( t in 3) can be used to represent the current time (or called the current moment); t 0 in formula (3) can be used to represent the earliest click behavior of the target business object for a certain business item; in formula (3)
  • the e ⁇ is an exponential function, and quality i can be used to characterize the ranking factor of a target business object for a certain business item.
  • the business output status and business conversion status of N delivered business data are respectively established through the business object set, and the unconverted browsing behavior characteristics of the target business object for unconverted business data are established, and the unconverted browsing behavior characteristics are established according to the unconverted browsing behavior.
  • Behavioral characteristics determine the sorting factor corresponding to the target business object and the unconverted item (that is, the business item included in the unconverted business data); then, sort the business data containing the unconverted item according to the sorting factor, and sort Select the target business data to deliver to the target business object.
  • the target business object did not convert the untransformed business data, but because the business output status of the untransformed business data has already been output, it can still be indicated
  • the target business object is interested in the business items in the unconverted business data; that is to say, the embodiment of this application can determine the target business object's real-time feedback data on the delivered business data according to the target business object , is the untransformed business data that has been output and is in the untransformed state, and these unconverted business data can indicate the potential preferences of the target business object, sort and deliver the business data containing unconverted items according to preferences, so that the delivered business
  • the data is in line with the preferences of the target business object, so it can improve the accuracy and reduce the cost of data transmission, and because the delivered data is in line with the preferences of the target business object, it can increase the probability of conversion behavior of the target business object, thereby improving conversion rate.
  • the embodiments of the present application can improve the accuracy of service data placement
  • the target business data in the sequence business data can be used (the target business data can be complete sequence business data, or can be Part of the business data in the sequence business data) is delivered to the target business object within the target time period; during this delivery process, the computer device can obtain the browsing feedback of the target business object for each business data in the target business data Data, so that the computer device can perform subsequent processing based on the browsing feedback data (for example, stop the delivery of sequential business data; or perform similar business data recommendation processing, etc.).
  • the browsing feedback data for example, stop the delivery of sequential business data; or perform similar business data recommendation processing, etc.
  • FIG. 5 is a schematic flow chart of recommending similar business data based on browsing feedback data of a target business object provided by an embodiment of the present application. As shown in FIG. 5 , the process may at least include the following steps S501 and S502 , and each step will be described separately below.
  • the business user corresponding to the target business object can click to trigger an operation to play and watch, and generate conversion behavior after playing and watching, then this At this time, the browsing feedback data can be converted feedback behavior data; correspondingly, the business user corresponding to the target business object can also click on the negative feedback control (such as close control, stop control, uninterested control and advertisement complaint control, etc.).
  • the negative feedback control such as close control, stop control, uninterested control and advertisement complaint control, etc.
  • the browsing feedback data can be negative feedback behavior data at this time; correspondingly, after the business user corresponding to the target business object plays and watches a certain business data, then that is After browsing and viewing the business data, the computer equipment can also count the target browsing time of the business user for the business data within the target time period, that is to say, the browsing feedback data can be the target business object for a certain business data.
  • Target browsing time the browsing feedback data includes at least one of the following: conversion feedback behavior data of the target business object for the target business data, negative feedback behavior data of the target business object for the target business data, and target browsing time of the target business object for the target business data.
  • the process for the computer device to recommend similar business data to the target business object includes: the computer device from the target business data, Obtain the conversion business data corresponding to the conversion feedback behavior data; obtain the converted items included in the conversion business data, and obtain the first similar item of the converted item; and use the business data of the first similar item as the target similar business data, and set the target Similar business data is delivered to the target business object.
  • the target business data is delivered, if the target business object has a conversion behavior on the business data A containing the business item "whitening mask”, then the business item "whitening mask” can be called a conversion item, and at this time it can be indicated
  • the target business object is interested in the business item "whitening mask”, so at this time, the computer device can deliver similar items (such as whitening essence and whitening cream) to the target business object according to the preferences of the target business object. ) corresponding business data.
  • the process for the computer device to recommend similar business data to the target business object includes: , to obtain the negative feedback business data corresponding to the negative feedback behavior data; and to obtain the negative feedback items included in the negative feedback business data, to obtain the second similar items of the negative feedback items;
  • Business data is used as business data with similar targets.
  • the business data with similar targets is filtered.
  • the target business object After the target business data is released, if the target business object generates negative feedback behavior data on the business data A containing the business item "whitening mask”, then the business item "whitening mask” can be called a negative feedback item , at this time, it can be indicated that the target business object has no interest in the item "whitening mask”, so at this time, the computer device can carry out a new round of delivery to the target business object according to the preferences of the target business object (within the target time period When launching in the next time period), business data corresponding to similar items (items of the same type, such as whitening essence, whitening cream, etc.) will no longer be delivered to the target business object.
  • the process for the computer device to recommend similar business data to the target business object includes: historical average browsing time; when the target browsing time is longer than the historical average browsing time, determine the business item included in the target business data as a potential conversion item, and use the business data of the third similar item of the potential conversion item as the target similar business data, and delivering target-similar business data to the target business object; wherein, during the process of delivering the target business data to the target business object, the computer device can obtain the target browsing time of the target business object for the target business data.
  • target business object's target browsing time for a certain business data in the target business data is greater than the target business object's historical average browsing time for the business data, it can indicate that during the delivery process, the target business object is The interest of business data increases, and in a new round of delivery, business data containing similar items of the same type can be delivered to target business objects.
  • the target business object's preferences can be determined based on the real-time feedback data (such as browsing feedback data) of the sequence business data by the target business object to perform corresponding business processing.
  • the target business object has a transformation behavior, it can continue to deliver the same type of business data to the target business object; when the target business object has a negative feedback behavior, it can stop delivering serial business data in time, and no longer Deliver the same type of business data to the target business object; when the browsing time of the target business object increases, you can continue to deliver the same type of business data to the target business object. That is to say, through the real-time feedback data of the target business object, the accuracy of delivering business data to the target business object can be improved, the conversion rate corresponding to the target business object can be improved, and the user experience can be improved.
  • FIG. 6 is a system structure diagram provided by an embodiment of the present application.
  • the system structure may include an advertisement content identification module, a feature determination module, a similar advertisement determination module, a click-through rate, a conversion rate estimation module, a ranking factor calculation module, an advertisement delivery module, a feedback data collection module, a behavior Time filter module and negative filter module.
  • the advertisement content recognition module can be used to carry out at least one of semantic content recognition and image content recognition on the advertisement data.
  • the feature determination module can be used to determine the target business object (actually, it can refer to the business object corresponding to the business user who has clicked but not converted, and can also be called the business object corresponding to the clicked unconverted user) for browsing an unconverted item Behavioral characteristics.
  • the Similar Ads Determination module can be used to obtain data for all ads containing the same unconverted item.
  • the click-through rate and estimated amount estimation module can be used to estimate the click-through rate and conversion rate of each advertisement data.
  • the ranking factor calculation module can be used to determine the ranking factor corresponding to a certain target business object and a certain unconverted item.
  • the advertising delivery module can sort the advertising data based on the sorting factors calculated by the sorting factor calculation module, and obtain part of the advertising data according to the needs of business scenarios for delivery.
  • the feedback data collection module may be used to collect browsing feedback data (for example, click behavior data and conversion behavior data, etc.) of each business object for each business data. It should be understood that the embodiments of the present application can determine users who have not been converted by clicking and the unconverted advertisement data corresponding to users who have not been converted by clicking based on the browsing feedback data of the business object.
  • the time window of the user's click unconverted behavior can be extracted; for example, if the click unconverted behavior in the last T days is taken, that is, assuming that the current moment is recorded as now, then the time window for a certain time period from now-T to now can be extracted. If a business object that has clicked but not converted on delivered advertisement data, the business object can be used as the target business object, and the delivered advertisement data can be used as the unconverted business data.
  • the negative filtering module can be used to filter the advertising data that the business object generates negative feedback behavior data.
  • Each module in FIG. 6 can be used to implement the processes corresponding to the above-mentioned FIGS. 3-5 .
  • For the implementation of each module refer to the description corresponding to the above-mentioned FIGS. 3-5 , which will not be repeated here.
  • FIG. 7 is a schematic structural diagram of a service data processing device provided in an embodiment of the present application.
  • the business data processing device can be a computer program (including program code) running in electronic equipment such as computer equipment, for example, the business data processing device is an application software; the business data processing device can be used to execute the method.
  • the business data processing device 71 may include: a log acquisition module 11 , a feature creation module 12 , a sorting factor determination module 13 , a data sorting module 14 , and a data delivery module 15 .
  • the log acquisition module 11 is configured to obtain the delivery log corresponding to the N delivered business data; the delivery log includes the business output status and the business conversion status of the business object set respectively for the N delivered business data; N is a positive integer ;
  • the business conversion state includes an untransformed state; the business output state includes an output state;
  • the feature establishment module 12 is configured to establish the unconverted browsing behavior of the target business object for the unconverted business data according to the business output status and the business conversion status of the N delivered business data respectively according to the business object set
  • the service output status of the target business object for the untransformed business data is the output status
  • the business conversion status is the untransformed status
  • the N delivered business data include the The untransformed business data
  • the business object set includes the target business object
  • the ranking factor determining module 13 is configured to determine the ranking factor corresponding to the target business object and the unconverted item according to the unconverted browsing behavior characteristics; the unconverted business data includes the unconverted item.
  • the data sorting module 14 is configured to sort the business data containing the unconverted items according to the sorting factor to obtain sequence business data;
  • the data delivery module 15 is configured to select target business data from the sequence of business data in a sorted order, and deliver the target business data to the target business object.
  • step S101-step S103 the implementation of the log acquisition module 11, feature establishment module 12, sorting factor determination module 13, data sorting module 14, and data delivery module 15 can refer to the description of step S101-step S103 in the embodiment corresponding to FIG. 3 above, No further details will be given here.
  • the delivery log further includes the record time stamps of the business object sets for the N delivered service data respectively;
  • the business object set includes a business object M i , and the N delivered service data
  • the business data includes the delivered business data G b , and the record time stamp of the business object Mi for the delivered business data G b refers to the time stamp of the business object Mi for the delivered business data G b
  • the operation record time stamp of the state change operation, the state change operation is used to indicate that the business output state of the business object M i for the delivered business data G b is converted from the non-output state to the output state; i and b are both positive integers;
  • the feature building module 12 includes: a data acquisition unit 121 and a feature determination unit 122 .
  • the data obtaining unit 121 is configured to obtain the object identifier of the business object M i and the service identifier of the delivered service data G b ;
  • the feature determination unit 122 is configured to use the object identifier of the business object M i , the service identifier of the delivered service data G b , the service object M i for the delivered service data G b
  • the formed data group is determined as the browsing behavior characteristic of the business object M i for the delivered business data G b ;
  • the characteristic determination unit 122 is further configured to determine the unconverted browsing behavior of the target business object for the unconverted business data in the browsing behavior characteristics of the business object set for the N delivered business data Behavioral characteristics.
  • step S102 For the implementation manners of the data acquisition unit 121 and the feature determination unit 122, reference may be made to the description of step S102 corresponding to FIG. 3 above, which will not be repeated here.
  • the feature determining unit 122 is further configured to set the service output status as the output status in the browsing behavior features of the business object set for the N delivered service data , and the browsing behavior characteristics in which the business conversion state is the non-transformed state are determined as the non-transformed browsing behavior characteristics;
  • the characteristic determining unit 122 is further configured to determine the business object corresponding to the object identifier included in the unconverted browsing behavior characteristic as the target business object, and the object identifier containing the target business object In the unconverted browsing behavior characteristics, the delivered business data corresponding to the business identifier included is determined as the unconverted business data targeted by the target business object, so as to obtain the targeted business data of the target business object.
  • the non-converted browsing behavior characteristics of the non-converted business data is further configured to determine the business object corresponding to the object identifier included in the unconverted browsing behavior characteristic as the target business object, and the object identifier containing the target business object In the unconverted browsing behavior characteristics, the delivered business data corresponding to the business identifier included is determined as the unconverted business data targeted by the target business object, so as to obtain the targeted business data of the target business object.
  • the non-converted browsing behavior characteristics of the non-converted business data is further configured to determine the business object corresponding to the object identifier included in the uncon
  • the sorting factor determining module 13 may include: a data identifying unit 131 , an identifier replacing unit 132 , and a sorting factor determining unit 133 .
  • the data identification unit 131 is configured to perform business identification on unconverted business data, and obtain a mapping relationship between unconverted business data and unconverted items; the mapping relationship is used to indicate that unconverted business data includes unconverted items;
  • the identifier replacement unit 132 is configured to obtain the item identifier of the unconverted item, and replace the service identifier of the unconverted business data contained in the unconverted browsing behavior feature with the item identifier of the unconverted item , determining the replaced unconverted browsing behavior characteristics as the target browsing behavior characteristics of the target business object for the unconverted items;
  • the ranking factor determining unit 133 is configured to determine the ranking factor corresponding to the target business object and the unconverted item according to the target browsing behavior characteristics.
  • the sorting factor determination unit 133 is further configured to count the total number of features of the Q target browsing behavior features, and determine the total number of features as the target business object for the unconverted item P The total output times of x ;
  • the sorting factor determination unit 133 is further configured to obtain the earliest record timestamp from the record timestamps respectively included in the Q target browsing behavior characteristics;
  • the sorting factor determining unit 133 is further configured to determine the sorting factor corresponding to the target business object and the unconverted item P x according to the current timestamp, the total output times, and the earliest record timestamp.
  • the business data containing the unconverted items includes Q target unconverted business data; the Q target unconverted business data all contain the same unconverted item Px; Q is less than or equal to N A positive integer of , x is a positive integer;
  • the data sorting module 14 may include: a delivery conversion rate determining unit 141 and a sorting unit 142 .
  • the delivery conversion rate determination unit 141 is configured to acquire the total delivery amount of the target unconverted business data R s for the set of business objects, the target unconverted business data R s being any of the Q target unconverted business data one;
  • the delivery conversion rate determining unit 141 is further configured to determine the target unconverted business data R s corresponding to Real-time delivery conversion;
  • the sorting unit 142 is configured to sort the business data containing the unconverted item P x according to the sorting factor and the real-time delivery conversion rates corresponding to the Q target unconverted business data, to obtain the The sequence business data described above.
  • step S103 for the implementation of delivery conversion rate determination unit 141 and sorting unit 142, reference may be made to the description of step S103 corresponding to FIG. 3 above, which will not be repeated here.
  • the sorting unit 142 is further configured to obtain the estimated delivery conversion rate corresponding to the business data to be delivered; the business data containing the unconverted item also includes the unconverted item P x Business data to be delivered;
  • the sorting unit 142 is further configured to, according to the order of magnitude between the real-time delivery conversion rates corresponding to the Q target unconverted business data and the estimated delivery conversion rates corresponding to the service data to be delivered, Sorting the business data containing the unconverted item P x to obtain initial sequence business data;
  • the sorting unit 142 is further configured to sequentially obtain K business data with the largest estimated delivery conversion rate from the initial sequence of business data; K is a positive integer less than W; W is the unconverted value of the Q targets The total quantity of business data and the business data to be delivered including the unconverted item P x ;
  • the sorting unit 142 is further configured to obtain the expected conversion unit investment resources, estimated output rate, and estimated conversion rate corresponding to the K pieces of business data;
  • the sorting unit 142 is further configured to sort the K business data according to the expected conversion unit investment resources, the estimated output rate, and the estimated conversion rate corresponding to the K business data respectively.
  • the business data is sorted to obtain the sequence of business data.
  • the business data R t is any one of the K business data, and t is a positive integer;
  • the sorting unit 142 is further configured to determine the initial delivery corresponding to the business data R t according to the expected conversion unit delivery resources corresponding to the business data R t , the estimated output rate, and the estimated conversion rate Units consume resources;
  • the sorting unit 142 is further configured to perform calculation processing on the sorting factor and the initial delivery unit consumption resource corresponding to the target service data R t to obtain the real-time delivery unit consumption resource corresponding to the target service data R t ;
  • the sorting unit 142 is further configured to sort the K pieces of business data according to the size sequence among the resources consumed by the K real-time delivery units, to obtain the sequence of business data.
  • the business data containing the untransformed items also includes unexported business data;
  • the business output status of the target business object for the unexported business data is the unexported status;
  • the sorting unit 142 is further configured to acquire the expected conversion unit investment resources, the estimated output rate, and the estimated conversion rate corresponding to the unoutputted business data;
  • the sorting unit 142 is further configured to determine the expected conversion unit investment resource corresponding to the unoutput business data, the estimated output rate, and the estimated conversion rate to determine the corresponding value of the unoutput business data.
  • the initial delivery unit consumes resources;
  • the sorting unit 142 is further configured to sort the K business data and the unoutput business data according to the size order between the resources consumed by the initial delivery unit corresponding to the unoutput business data and the resources consumed by the K real-time delivery units, to obtain the sequence business data.
  • the business data processing device 71 may further include: a feedback data acquisition module 16 and a data recommendation module 17 .
  • the feedback data acquisition module 16 is configured to acquire the browsing feedback data of the target business object for the target business data within the target time period
  • the data recommendation module 17 is configured to deliver similar business data to the target business object according to the browsing feedback data.
  • step S501-step S502 corresponding to FIG. 5 above, which will not be repeated here.
  • the browsing feedback data includes conversion feedback behavior data of the target business object for the target business data
  • the data recommendation module 17 may include: a converted item acquisition unit 171 and a first data recommendation unit 172 .
  • the converted item acquiring unit 171 is configured to acquire the converted business data corresponding to the conversion feedback behavior data from the target business data;
  • the conversion item acquiring unit 171 is further configured to acquire the conversion item included in the conversion business data, and acquire the first similar item of the conversion item;
  • the first data recommendation unit 172 is configured to use the business data to which the first similar item belongs as the target similar business data, and deliver the target similar business data to the target business object.
  • step S502 For the implementation of the conversion item acquisition unit 171 and the first data recommendation unit 172, reference may be made to the description in step S502 corresponding to FIG. 5 above, which will not be repeated here.
  • the browsing feedback data includes the negative feedback behavior data of the target business object for the target business data
  • the data recommendation module 17 may include: a negative feedback item acquisition unit 173 and a second data recommendation unit 174 .
  • the negative feedback item acquisition unit 173 is configured to acquire the negative feedback business data corresponding to the negative feedback behavior data from the target business data;
  • the negative feedback item acquiring unit 173 is further configured to acquire the negative feedback item included in the negative feedback business data, and acquire the second similar item of the negative feedback item;
  • the second data recommending unit 174 is configured to use the business data to which the second similar item belongs as the target similar business data, and filter the target similar business data during the delivery process of the business data for the target business object.
  • the implementation of the negative feedback item acquisition unit 173 and the second data recommendation unit 174 can refer to the description in the step S502 corresponding to FIG. 5 above, which will not be repeated here.
  • the browsing feedback data includes the browsing time of the target business object for the target business data
  • the data recommendation module 17 may include: a duration acquisition unit 175 and a third data recommendation unit 176 .
  • the duration acquisition unit 175 is configured to acquire the historical average browsing duration of the target business object for the target business data
  • the third data recommendation unit 176 is configured to determine the business item included in the target business data as a potential conversion item when the target browsing time is longer than the historical average browsing time, and use the business data to which the third similar item of the potential conversion item belongs as the target Similar business data, deliver target similar business data to target business objects.
  • step S502 For the implementation of the duration acquiring unit 175 and the third data recommending unit 176, refer to the description in step S502 corresponding to FIG. 5 above, which will not be repeated here.
  • FIG. 8 is a schematic structural diagram of a computer device provided by an embodiment of the present application.
  • the business data processing apparatus 71 in the embodiment corresponding to the above-mentioned FIG. 7 can be applied to the above-mentioned computer equipment 8000 (referred to as electronic equipment), and the above-mentioned computer equipment 8000 can include: a processor 8001, a network interface 8004 and a memory 8005.
  • the above computer device 8000 further includes: a user interface 8003 and at least one communication bus 8002. Among them, the communication bus 8002 is used to realize connection and communication between these components.
  • the user interface 8003 may include a display screen (Display) and a keyboard (Keyboard), and the user interface 8003 may also include a standard wired interface and a wireless interface.
  • the network interface 8004 may include a standard wired interface and a wireless interface (such as a WI-FI interface).
  • the memory 8005 can be a high-speed RAM memory, or a non-volatile memory (Non-Volatile Memory), such as at least one disk memory.
  • the memory 8005 may also be at least one storage device located away from the aforementioned processor 8001 .
  • the memory 8005 as a computer-readable storage medium may include an operating system, a network communication module, a user interface module, and a device control application program.
  • the network interface 8004 can provide a network communication function; the user interface 8003 is mainly used to provide an input interface for the user; and the processor 8001 can be used to call the device control application stored in the memory 8005 program to implement the business data processing method provided in the embodiment of this application.
  • the computer equipment 8000 described in the embodiment of this application can execute the description of the business data processing method corresponding to Figure 3 to Figure 5, and can also execute the description of the business data processing device 71 corresponding to Figure 7, here No longer. In addition, the description of the beneficial effect of adopting the same method will not be repeated here.
  • the embodiment of the present application also provides a computer-readable storage medium, and the above-mentioned computer-readable storage medium stores the computer program executed by the computer device 8000, and the above-mentioned computer program includes program instructions, when the above-mentioned processor executes the above-mentioned program instructions , the service data processing method corresponding to FIG. 3 to FIG. 5 can be executed, so details will not be repeated here. In addition, the description of the beneficial effect of adopting the same method will not be repeated here.
  • the above-mentioned computer-readable storage medium may be the service data processing apparatus provided in the embodiment of the present application or the internal storage unit of the computer equipment, such as the hard disk or memory of the computer equipment.
  • the computer-readable storage medium can also be an external storage device of the computer device, such as a plug-in hard disk equipped on the computer device, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, Flash card (Flash Card), etc.
  • the computer-readable storage medium may also include both an internal storage unit of the computer device and an external storage device.
  • the computer-readable storage medium is used to store the computer program and other programs and data required by the computer device.
  • the computer-readable storage medium can also be used to temporarily store data that has been output or will be output.
  • An embodiment of the present application provides a computer program product or computer program, where the computer program product or computer program includes computer instructions, and the computer instructions are stored in a computer-readable storage medium.
  • the processor of the computer device reads the computer instruction from the computer-readable storage medium, and the processor executes the computer instruction, so that the computer device executes the business data processing method provided by the embodiment of the present application.
  • These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing apparatus to operate in a specific manner, such that the instructions stored in the computer-readable memory produce an article of manufacture comprising instruction means, the instructions
  • the device implements the functions specified in one or more blocks of the flowchart and/or one or more blocks of the structural schematic diagram.
  • These computer program instructions can also be loaded onto a computer or other programmable data processing device, causing a series of operational steps to be performed on the computer or other programmable device to produce a computer-implemented process, thereby
  • the instructions provide steps for implementing the functions specified in one or more steps of the flowchart and/or one or more blocks in the structural illustration.

Landscapes

  • Business, Economics & Management (AREA)
  • Strategic Management (AREA)
  • Engineering & Computer Science (AREA)
  • Accounting & Taxation (AREA)
  • Development Economics (AREA)
  • Finance (AREA)
  • Physics & Mathematics (AREA)
  • Economics (AREA)
  • Game Theory and Decision Science (AREA)
  • Marketing (AREA)
  • Entrepreneurship & Innovation (AREA)
  • General Business, Economics & Management (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Probability & Statistics with Applications (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

一种业务数据处理方法、装置、电子设备、计算机可读存储介质及计算机程序产品，业务数据处理方法包括：获取N个已投放业务数据对应的投放日志；根据投放日志中业务对象集合分别针对N个已投放业务数据的业务输出状态与业务转化状态，建立目标业务对象针对未转化业务数据的未转化浏览行为特征；目标业务对象针对未转化业务数据的业务输出状态为已输出状态且业务转化状态为未转化状态；根据未转化浏览行为特征确定目标业务对象与未转化物品共同对应的排序因子，根据排序因子对包含有未转化物品的业务数据进行排序，得到序列业务数据，按照排序顺序从序列业务数据选择目标业务数据，将目标业务数据投放至目标业务对象。

Description

一种业务数据处理方法、装置、电子设备、计算机可读存储介质及计算机程序产品
相关申请的交叉引用
本申请基于申请号为202110919651.0、申请日为2021年08月11日的中国专利申请提出，并要求该中国专利申请的优先权，该中国专利申请的全部内容在此引入本申请作为参考。
技术领域
本申请涉及计算机技术领域，尤其涉及一种业务数据处理方法、装置、电子设备、计算机可读存储介质及计算机程序产品。
背景技术
随着计算机技术的飞速发展，越来越多的用户通过电子设备上的应用发布或者获取信息，创建业务数据的用户可以通过电子设备上投放业务数据的应用来实现业务数据的投放。
在通常情况下，应用平台在将业务数据投放至用户群体前，会先对各个业务数据进行排序，再将排序后的业务数据按序投放至用户群体。而为了提高转化率(也就是通过业务数据所产生下单、购买等转化行为的用户数量，与点击业务数据的用户数量之间的比值)，相关技术会重复多次且始终按照同样的顺序对全部的业务数据进行投放，然而，这种方式会造成业务数据的过度投放，对平台的资源(计算资源如处理器线程线程、通信资源如带宽)造成了较大的浪费，如此，不仅增加了数据处理成本，还难以改善业务数据的投放准确度。
发明内容
本申请实施例提供一种业务数据处理方法、装置、电子设备、计算机可读存储介质及计算机程序产品，可以提高业务数据的投放精准度，提高业务数据的投放准确度，节约资源。
本申请实施例提供了一种业务数据处理方法，包括：
获取N个已投放业务数据对应的投放日志；投放日志包括业务对象集合分别针对N个已投放业务数据的业务输出状态与业务转化状态；N为正整数；业务转化状态包括未转化状态；业务输出状态包括已输出状态；
根据业务对象集合分别针对N个已投放业务数据的业务输出状态与业务转化状态，建立目标业务对象针对未转化业务数据的未转化浏览行为特征；目标业务对象针对未转化业务数据的业务输出状态为已输出状态且业务转化状态为未转化状态；N个已投放业务数据包括未转化业务数据，业务对象集合包括目标业务对象；
根据未转化浏览行为特征，确定目标业务对象与未转化物品共同对应的排序因子，根据目标业务对象与未转化物品共同对应的排序因子，对包含有未转化物品的业务数据进行排序，得到序列业务数据，按照排序顺序从序列业务数据选择目标业务数据投放至 目标业务对象；未转化业务数据中包含未转化物品。
本申请实施例提供了一种业务数据处理装置，包括：
日志获取模块，用于获取N个已投放业务数据对应的投放日志；投放日志包括业务对象集合分别针对N个已投放业务数据的业务输出状态与业务转化状态；N为正整数；业务转化状态包括未转化状态；业务输出状态包括已输出状态；
特征建立模块，用于根据业务对象集合分别针对N个已投放业务数据的业务输出状态与业务转化状态，建立目标业务对象针对未转化业务数据的未转化浏览行为特征；目标业务对象针对未转化业务数据的业务输出状态为已输出状态且业务转化状态为未转化状态；N个已投放业务数据包括未转化业务数据，业务对象集合包括目标业务对象；
排序因子确定模块，用于根据未转化浏览行为特征，确定目标业务对象与未转化物品共同对应的排序因子；未转化业务数据中包含未转化物品。
数据排序模块，用于根据目标业务对象与未转化物品共同对应的排序因子，对包含有未转化物品的业务数据进行排序，得到序列业务数据；
数据投放模块，用于按照排序顺序从序列业务数据选择目标业务数据投放至目标业务对象。
本申请实施例提供了一种电子设备，包括：处理器和存储器；
存储器存储有计算机程序，计算机程序被处理器执行时，使得处理器执行本申请实施例中的方法。
本申请实施例提供了一种计算机可读存储介质，计算机可读存储介质存储有计算机程序，计算机程序包括程序指令，程序指令当被处理器执行时，执行本申请实施例中的方法。
本申请实施例提供了一种计算机程序产品或计算机程序，该计算机程序产品或计算机程序包括计算机指令，该计算机指令存储在计算机可读存储介质中，电子设备的处理器从计算机可读存储介质读取该计算机指令，处理器执行该计算机指令，使得该电子设备执行本申请实施例中提供的方法。
在本申请实施例中，根据目标业务对象对已投放业务数据的实时反馈数据(业务输出状态、业务转化状态等)，来确定出目标业务对象对应的为已输出状态且为未转化状态的未转化业务数据，由于目标业务对象对未转化业务数据的业务输出状态为已输出状态，表明该目标业务对象是对该未转化业务数据中的业务物品存在兴趣；从而，通过计算出该目标业务对象与该业务物品共同对应的排序因子，再根据该排序因子将包含相同未转化物品的业务数据进行排序，并按照排序顺序从序列业务数据中选择目标业务数据投放至目标业务对象的过程中，按照这些未转化业务数据可表明目标业务对象的潜在喜好(目标业务对象对未转化业务数据中的未转化物品有较高概率会发生转化)，按照喜好将包含未转化物品的业务数据进行排序并投放，使得投放的业务数据是符合目标业务对象的喜好的，所以更为精准，减少了无效投放，故能够可以节约处理资源。
附图说明
为了更清楚地说明本申请实施例中的技术方案，下面将对本申请实施例中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本申请的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其他的附图。
图1是本申请实施例提供的一种网络架构图；
图2a-图2c是本申请实施例提供的一种确定排序因子的场景示意图；
图3是本申请实施例提供的一种业务数据处理方法的流程示意图；
图4是本申请实施例提供的另一种业务数据处理方法的流程示意图；
图5是本申请实施例提供的一种基于目标业务对象的浏览反馈数据，进行相似业务数据推荐处理的流程示意图；
图6是本申请实施例提供的一种***结构图；
图7是本申请实施例提供的一种业务数据处理装置的结构示意图；
图8是本申请实施例提供的一种计算机设备的结构示意图。
具体实施方式
下面将结合本申请实施例中的附图，对本申请实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例仅仅是本申请一部分实施例，而不是全部的实施例。基于本申请中的实施例，本领域普通技术人员在没有作出创造性劳动前提下所获得的所有其他实施例，都属于本申请保护的范围。
请参见图1，图1是本申请实施例提供的一种网络架构图。如图1所示，该网络架构可以包括业务服务器1000和用户终端集群，用户终端集群可以包括一个或者多个用户终端，这里将不对用户终端的数量进行限制。如图1所示，多个用户终端可以包括用户终端100a、用户终端100b、用户终端100c、…、用户终端100n；如图1所示，用户终端100a、用户终端100b、用户终端100c、…、用户终端100n可以分别与业务服务器1000进行网络连接，以便于每个用户终端可以通过该网络连接与业务服务器1000之间进行数据交互。
可以理解的是，如图1所示的每个终端设备(称为用户终端)均可以安装有目标应用，当该目标应用运行于各终端设备中时，可以分别与图1所示的业务服务器1000之间进行数据交互，使得业务服务器1000可以接收来自于每个终端设备的业务数据。其中，该目标应用可以包括具有显示文字、图像、音频以及视频等数据信息功能的应用。如，应用可以为支持业务数据投放的应用，如应用可以为多媒体类应用(例如，视频应用)、娱乐类应用(例如，游戏应用)和社交应用等。应当理解，本申请实施例中的业务数据可以为媒体数据(如，广告数据)，以下将以业务数据为广告数据为例进行阐述。
本申请实施例中的业务服务器1000可以根据这些应用从业务创建对象(业务创建对象可以是指创建广告数据的用户在应用中的应用账号；业务创建对象也可以是指创建广告数据的用户对应的终端设备；其中，创建广告数据的用户也可称为业务创建用户或广告主)处，获取到业务创建用户所创建的待进行投放的广告数据；本申请实施例中的业务服务器1000还可以根据这些应用从业务投放平台(比如，当业务数据为广告数据时，该业务投放平台也就是广告投放平台)对应的账号或终端设备处，获取到这些被投放的广告数据的投放日志(广告日志，也就是广告数据的相关投放数据)，每个被投放的广告数据均可以对应有一个投放日志(也可以一个投放日志中包含有所有被投放的广告数据的投放相关数据)，投放日志中可包括每个被投放的广告数据的曝光量(将广告数据投放至用户群体后、点击并播放了该广告数据的业务对象(也就是被投放广告数据的用户在应用中的应用账号)的对象标识的数量)和实时转化数(将广告数据投放至用户群体后、点击并转化了该广告数据的业务对象的对象标识的数量)。
应当理解，被投放的广告数据可称为已投放业务数据(已投放广告数据)，以下将已投放业务数据称之为已投放广告数据进行阐述。在业务服务器1000获取到已投放广告数据的广告日志后，也就可以获取到被投放的业务对象集合中，每个业务对象点击播放的已投放广告数据、也可以确定每个业务对象产生转化行为(比如，收藏、下单购买 和到店访问等)的已投放广告数据。当某个业务对象点击播放了某个已投放业务数据、但未产生转化行为时，该业务对象可称为目标业务对象、该已投放业务数据可称为未转化广告数据(也可称为未转化业务数据)。随后，业务服务器1000可以建立目标业务对象针对未转化广告数据的未转化浏览行为特征，并根据该未转化浏览行为特征确定该目标业务对象与未转化物品(也就是未转化广告数据中包括的物品，可以是实际物品，例如视频、电子产品；可以是虚拟物品，例如游戏道具，各个广告数据中可以包括不同待推广的物品，业务对象可通过广告数据对物品进行下单和购买等转化)共同对应的排序因子。随后，可基于该排序因子对包含有相同未转化物品的业务数据(广告数据)进行排序，再按序在这些排序后的业务数据中选择目标业务数据(目标广告数据)投放至目标业务对象。其中，对于建立目标业务对象针对未转化广告数据的未转化浏览行为特征，并根据该未转化浏览行为特征确定该目标业务对象与未转化物品共同对应的排序因子，及其将目标业务数据投放业务数据至目标业务对象的具体实现方式，可以参见后续图3所对应的描述。
应当理解，通过业务对象对已投放广告数据的实时反馈数据，可以确定某个业务对象对哪些已投放广告数据进行了点击观看(也就是点击播放)，由此可以确定业务对象对物品的喜好；从而可以根据业务对象对物品的喜好，来为业务对象投放包含相同物品的广告数据，因为这些广告数据是符合业务对象的喜好的，当包含相同物品的广告数据被投放至业务对象后，业务对象很可能会对所投放的广告数据进行点击观看并产生转化行为，从而可以提高广告数据的转化率。
本申请实施例可以在多个终端设备中选择一个终端设备作为目标终端设备(该目标终端设备可以为业务创建用户对应的终端设备，也可以为业务投放平台对应的终端设备)，该终端设备可以包括：智能手机、平板电脑、笔记本电脑、桌上型电脑、智能电视、智能音箱、台式计算机、智能手表和车载设备等携带多媒体数据处理功能(例如，视频数据播放功能、音乐数据播放功能)的智能终端，但并不局限于此。例如，本申请实施例可以将图1所示的用户终端100a作为该目标终端设备，该目标终端设备中可以集成有上述目标应用，此时，该目标终端设备可以通过该目标应用与业务服务器1000进行数据交互。
可以理解的是，本申请实施例提供的业务数据处理方法可以由计算机设备(称为电子设备)执行，计算机设备包括但不限于终端设备或业务服务器。其中，业务服务器可以是独立的物理服务器，也可以是多个物理服务器构成的服务器集群或者分布式***，还可以是提供云服务、云数据库、云计算、云函数、云存储、网络服务、云通信、中间件服务、域名服务、安全服务、内容分发网络(CDN，Content Delivery Network)、以及大数据和人工智能平台等基础云计算服务的云服务器。
其中，终端设备以及业务服务器可以通过有线或无线通信方式进行直接或间接地连接，本申请实施例在此不做限制。
可以理解的是，上述计算机设备(如上述业务服务器1000、用户终端100a和用户终端100b等)可以是一个分布式***中的一个节点，其中，该分布式***可以为区块链***，该区块链***可以是由该多个节点通过网络通信的形式连接形成的分布式***。其中，节点之间可以组成点对点(P2P，Peer To Peer)网络，点对点网络对应的P2P协议是一个运行在传输控制协议(TCP，Transmission Control Protocol)之上的应用层协议。在分布式***中，任意形式的计算机设备，比如业务服务器和终端设备等电子设备都可以通过加入该点对点网络而成为该区块链***中的一个节点。为便于理解，以下将对区块链的概念进行说明。
区块链是一种采用分布式数据存储、点对点传输、共识机制以及加密算法等计算机 技术的新型应用模式，用于对数据按时间顺序进行整理，并加密成账本，使其不可被篡改和伪造，同时可进行数据的验证、存储和更新。当计算机设备为区块链节点时，由于区块链的不可被篡改特性与防伪造特性，可以使得本申请实施例中的数据(如针对已投放业务数据的投放相关数据等)具备真实性与安全性，从而可以使得基于这些数据进行相关业务数据处理后得到的结果，相对于未结合区块链所获得的业务数据处理结果，更为可靠。
为便于理解，请参见图2a-图2c，图2a-图2c是本申请实施例提供的一种确定排序因子的场景示意图。其中，如图2a-图2b所示的用户终端100a-用户终端100e属于上述图1所对应的各个终端设备。
如图2a-图2c所示的场景是以向业务对象集合200投放广告数据集合20为例的投放场景，其中，该广告数据集合20中可包括广告数据2001、广告数据2002、广告数据2003、…、广告数据2009共9条广告数据；其中，广告数据2001、广告数据2002、广告数据2003、广告数据2004和广告数据2008中均包含有业务物品“粉底液”；广告数据2005和广告数据2006中均包含有业务物品“纯牛奶”；广告数据2007和广告数据2009中包含有业务物品“饼干”。因为将广告数据集合20向业务对象集合200投放后，该广告数据集合中的每个广告数据均为已经投放了的广告数据，故以下将每个广告数据均称为已投放广告数据。
如图2a所示，该业务对象集合200中可包括业务对象a1、业务对象a2、业务对象a3、业务对象a4以及业务对象a5，并且，可以获取到各个业务对象针对这些已投放广告数据的点击行为(点击行为即为业务对象点击播放了该广告数据)与转化行为(转化行为即为业务对象在点击播放某个业务数据后，产生了下单或购买行为(即下单或购买了业务数据中所推广的业务物品))；然后由此可以针对每个业务对象，确定出存在点击行为同时存在未转化行为的已投放广告数据。其中，在获取到各个业务对象针对已投放广告数据的点击行为与转化行为后，可以根据业务对象的对象标识(用于表征业务对象，该对象标识可以为对象名称、对象身份标识(ID，Identity Document)等)、已投放广告数据的广告标识(用于表征广告数据，广告标识可以为广告名称、广告ID等)、是否存在点击行为、是否存在转化行为、产生点击行为的记录时间等数据，建立一个业务对象与一个广告数据之间的数据组。
这里以业务对象a1与已投放广告数据2001为例进行说明。业务对象a1点击播放了已投放广告数据2001，但是未对已投放广告数据产生转化行为，则针对该业务对象a1与已投放广告数据2001的数据组可为[a1，2001，1，0，14:00]，其中，该数据组中的第三个数值1可用于表征该业务对象a1对该已投放广告数据2001产生了点击行为，该数据组中的第四个数值0可用于表征业务对象a1对该已投放广告数据2001未产生转化行为，该数据组中的时间信息14:00可为2021年8月7日中的14:00，该时间信息可用于表征业务对象a1对该已投放广告数据2001产生点击行为的时间，也就是说，业务对象a1在2021年8月7日14:00点击播放了该已投放广告数据2001。应当理解，业务对象a1针对已投放广告数据是否存在点击行为可理解为业务对象a1针对已投放广告数据的业务输出状态，例如，业务对象a1针对已投放广告数据2001存在点击行为，即为业务对象a1点击播放了该已投放广告数据2001，则业务对象a1针对已投放广告数据2001的业务输出状态则为已输出状态；而若业务对象a1针对已投放广告数据2001不存在点击行为，即为业务对象并未点击播放已投放广告数据2001，该业务对象a1针对已投放广告数据2001的业务输出状态为未输出状态。业务对象a1针对已投放广告数据是否存在转化行为可以理解为业务对象针对已投放广告数据的业务转化状态，例如，若业务对象a1对已投放广告数据存在转化行为，则业务对象a1针对已投放广告数据的业务 转化状态即为已转化状态；而若业务对象a1对已投放广告数据不存在转化行为，则业务对象a1针对已投放广告数据的业务转化状态即为未转化状态。
应当理解，上述针对该业务对象a1与已投放广告数据2001的数据组[a1，2001，1，0，14:00]可以确定为是业务对象a1针对已投放广告数据2001的浏览行为特征。同理，可构建得到每个业务对象针对所有已投放广告数据的数据组，也就是说，可构建得到每个业务对象针对所有已投放广告数据的浏览行为特征。在得到每个业务对象针对所有已投放广告数据的浏览行为特征后，因为各个浏览行为特征中包含有各个业务对象的对象标识、已投放广告数据的广告标识、是否存在点击行为、是否存在转化行为、发生点击行为的记录时间(若并未发生点击行为，则该记录时间可为指定值，比如，空值)，那么通过各个浏览行为特征，即可针对每个业务对象确定出存在点击行为同时存在未转化行为的已投放广告数据。这些针对业务对象存在点击行为但未产生转化行为的已投放广告数据，可称为针对该业务对象的未转化业务数据(在图2a-图2c中将称之为未转化广告数据)。
这里以业务对象a1为例说明业务对象对应的未转化广告数据。通过包含有业务对象a1的对象标识的浏览行为特征，可以确定出业务对象a1点击播放了已投放广告数据2001、已投放广告数据2002、已投放广告数据2003、已投放广告数据2004、已投放广告数据2005和已投放广告数据2006，同时针对已投放广告数据2001、已投放广告数据2002、已投放广告数据2003、已投放广告数据2004、已投放广告数据2005和已投放广告数据2006，业务对象a1均未发生下单或购买等转化行为，所以针对业务对象a1，可以确定出已投放广告数据集合20a(包括已投放广告数据2001、已投放广告数据2002、已投放广告数据2003、已投放广告数据2004、已投放广告数据2005和已投放广告数据2006)为业务对象a1对应的未转化广告数据。
同理，可以在广告数据集合20中，确定出针对业务对象a2的未转化广告数据(如，已投放广告数据集合20b)、针对业务对象a3的未转化广告数据(如，已投放广告数据集合20c)、针对业务对象a4的未转化广告数据(如，已投放广告数据集合20d)、针对业务对象a5的未转化广告数据(如，已投放广告数据集合20e)。进一步地，可以将每个业务对象作为目标业务对象，确定出每个目标业务对象针对各自未转化广告数据的浏览行为特征。为便于理解，一个目标业务对象针对一个未转化广告数据的浏览行为特征可称之为未转化浏览行为特征。其中，图2a中，业务对象a3和业务对象a4，以及已投放广告数据集合20b、已投放广告数据集合20c和已投放广告数据集合20d等，是以省略号(…)的形式示出的。
如图2b所示，以目标业务对象为业务对象a1(以下将称为目标业务对象a1)为例(因为已投放广告数据集合20a为业务对象a1对应的未转化广告数据，则以下将会将已投放广告数据集合20a中的每个已投放广告数据称为未转化广告数据)说明目标业务对象针对未转化广告数据的未转化浏览行为特征。目标业务对象a1针对未转化广告数据2001的浏览行为特征为数据组[a1，2001，1，0，14:00]，该数据组[a1，2001，1，0，14:00]即可作为该目标业务对象a1针对未转化广告数据2001的未转化浏览行为特征2000a；同理，目标业务对象a1针对未转化广告数据2002的浏览行为特征为数据组[a1，2002，1，0，14:10]、目标业务对象a1针对未转化广告数据2003的浏览行为特征为数据组[a1，2003，1，0，14:20]、目标业务对象a1针对未转化广告数据2004的浏览行为特征为数据组[a1，2004，1，0，14:30]、目标业务对象a1针对未转化广告数据2005的浏览行为特征为数据组[a1，2005，1，0，14:40]、目标业务对象a1针对未转化广告数据2006的浏览行为特征为数据组[a1，2006，1，0，14:50](以上各个数据组中的时间信息可均处于2021年8月7日中)；由此可以确定，数据组[a1，2002，1，0，14:10]即 可作为目标业务对象a1针对未转化广告数据2002的未转化浏览行为特征2000b，数据组[a1，2003，1，0，14:10]即可作为未转化广告数据2003的未转化浏览行为特征2000c，数据组[a1，2004，1，0，14:10]即可作为针对未转化广告数据2004的未转化浏览行为特征2000d，数据组[a1，2005，1，0，14:10]即可作为针对未转化广告数据2005的未转化浏览行为特征2000e，数据组[a1，2006，1，0，14:10]即可作为针对未转化广告数据2006的未转化浏览行为特征2000f。
在本申请实施例中，可以对目标业务对象a1所对应的这些未转化广告数据分别进行广告识别，以确定出这些未转化广告数据分别包含的业务物品(因为广告数据为未转化广告，则为了理解，以下将会将未转化广告数据包含的业务物品称为未转化物品)，再将各个未转化浏览行为特征中包含的未转化广告数据的广告标识，替换为未转化广告数据包含的业务物品的物品标识，由此可得到目标业务对象a1针对各个业务物品的目标浏览行为特征。
这里以未转化广告数据为例对目标浏览行为特征进行说明。本申请实施例的示例性场景中，广告数据2001、广告数据2002、广告数据2003、广告数据2004和广告数据2008中均包含有业务物品“粉底液”，广告数据2005、广告数据2006、广告数据2007和广告数据2008中均包含有业务物品“纯牛奶”；从而通过广告识别，可以确定该未转化广告数据2001中包含未转化物品“粉底液”，如果“粉底液”的物品标识为FDY001，那么对于目标业务对象a1针对未转化广告数据2001的未转化浏览行为特征[a1，2001，1，0，14:00]，可将该数据组[a1，2001，1，0，14:00]中的广告标识2001替换为“粉底液”的物品标识FDY001，由此可得到目标业务对象a1针对未转化物品“粉底液”的目标浏览行为特征2000a’为[a1，FDY001，1，0，14:00]。同理，可以确定出目标业务对象a1针对未转化物品“粉底液”的目标浏览行为特征2000b’(即数据组[a1，FDY001，1，0，14:10]，对应于未转化浏览行为特征2000b)、针对未转化物品“粉底液”的目标浏览行为特征2000c’(即[a1，FDY001，1，0，14:20]，对应于未转化浏览行为特征2000c)、针对未转化物品“粉底液”的目标浏览行为特征2000d’(即[a1，FDY001，1，0，14:30]，对应于未转化浏览行为特征2000d)、针对未转化物品“纯牛奶”的目标浏览行为特征2000e’(即[a1，CN001，1，0，14:40]，对应于未转化浏览行为特征2000e，CN001为“纯牛奶”的物品标识)、针对未转化物品“纯牛奶”的目标浏览行为特征2000f’(即[a1，CN001，1，0，14:50]，对应于未转化浏览行为特征2000f)。
在本申请实施例中，可以在目标业务对象a1针对各个业务物品的目标浏览行为特征中，获取到属于同一个业务物品的目标浏览行为特征(即物品标识相同的目标浏览行为特征)。根据属于同一个业务物品的目标浏览行为特征，可以确定出目标业务对象a1与该业务物品共同对应的排序因子(排序因子可以用于表征目标业务对象对于某个业务物品的感兴趣程度；也就是说，目标业务对象与某个业务物品共同对应的排序因子可理解为该目标业务对象对这个业务物品的兴趣度。本申请实施例中一个目标业务对象与某个业务物品的排序因子与该目标业务对象对该业务物品的点击次数相关联，而点击次数可表明目标业务对象针对该业务物品的感兴趣程度，若目标业务对象针对某个业务物品的点击次数越多，则可表明目标业务对象对该业务物品的感兴趣程度也越高，那么目标业务对象与该业务物品共同对应的排序因子也会较大)。
以下将举例说明确定排序因子的过程。如2c所示，以未转化物品“粉底液”为例说明确定排序因子的过程；因为数据组[a1，2001，1，0，14:00]、数据组[a1，2002，1，0，14:10]、数据组[a1，2003，1，0，14:20]和数据组[a1，2004，1，0，14:30]均为目标业务对象a1针对未转化物品“粉底液”的浏览行为特征，从而可以根据这些浏览行为特征统计确定出目标业务对象a1针对未转化物品“粉底液”的总点击次数(一个浏览 行为特征对应于一次点击行为，那么该总点击次数即可为该浏览行为特征的总数量，即为4)；随后，可以在这些浏览行为特征中，获取到最早记录时间(也就是目标业务对象a1针对未转化物品“粉底液”，最早发生点击行为的时间，也可称为最早点击时间)，且该最早记录时间为14:00(即2021年8月7日14:00)；同时，也可以获取到当前时间(当前所处的时刻，可以理解为当前获取这些信息的时刻)；最后，根据总点击次数、最早记录时间以及当前时间，即可确定出目标业务对象a1与未转化物品“粉底液”共同对应的排序因子1。也就是说，目标业务对象a1与未转化物品“粉底液”共同对应的排序因子1，是与目标业务对象a1对该未转化物品“粉底液”的点击次数相关联的。其中，对于根据总点击次数、最早记录时间以及当前时间确定排序因子的具体实现方式，可以参见后续图3-图4所对应的描述。
同理，也可以确定出目标业务对象a1与未转化物品“纯牛奶”共同对应的排序因子2。在本申请实施例中，可以在广告数据集合20中，获取到包含有未转化物品“粉底液”与包含有未转化物品“纯牛奶”的已投放广告数据(即广告数据2001、广告数据2002、广告数据2003、广告数据2008、广告数据2005和广告数据2006。因为这些已投放广告数据在后续会进行排序，所以为便于区别理解，也可将这些包含有未转化物品“粉底液”与包含有未转化物品“纯牛奶”的已投放广告数据称为待排序已投放广告数据)。同时，也可以在还未投放至业务对象集合200的候选广告数据集合中，获取到包含有未转化物品“粉底液”与包含有未转化物品“纯牛奶”的待投放广告数据。其中，该候选广告数据集合可以是指等待投放的广告数据所组成的集合。可将这些包含有未转化物品“粉底液”与包含有未转化物品“纯牛奶”的待投放广告数据、以及包含有未转化物品“粉底液”与包含有未转化物品“纯牛奶”的已投放广告数据(即待排序已投放广告数据)，均确定为包含有未转化物品“粉底液”与包含有未转化物品“纯牛奶”的广告数据。
在本申请实施例中，也可以获取到目标业务对象a1针对广告数据集合20并未产生点击行为的已投放广告数据(因为业务对象a1并未针对这些广告数据进行点击播放，则目标业务对象a1并未播放观看这些广告数据，本申请实施例可将这些广告数据称为未播出广告数据、未点击广告数据或未输出广告数据)，若未播出广告数据为广告数据2009，则根据排序因子1与排序因子2，可对上述包含有未转化物品“粉底液”与包含有未转化物品“纯牛奶”的广告数据，以及未输出广告数据2009，进行排序，得到序列广告数据，并按照排序后的结果选择目标广告数据投放至目标业务对象a1。其中，对于排序方式以及选择目标广告数据的方式，可以参见后续图3所对应的描述。
应当理解，在本申请实施例中，通过获取目标业务对象a1针对各个已投放广告数据的点击行为、转化行为和发生点击行为的时间等实时反馈信息，可以确定出目标业务对象a1发生了点击行为但未发生转化行为的未转化广告数据。而通过建立目标业务对象a1针对各个未转化业务广告数据的未转化浏览行为特征，可以进而基于未转化浏览行为特征确定出目标业务a1针对未转化物品的排序因子。因为目标业务a1点击了这些未转化广告数据，即可表明该目标业务对象a1对这些广告数据对应的未转化物品是存在兴趣的，从而本申请实施例可基于未转化浏览行为特征确定出未转化物品的排序因子，再基于排序因子对包含有未转化物品的广告数据与未点击广告数据进行排序，由于排序因子的存在，那么在按照从大到小的顺序排序时，增加了将包含有未转化物品的广告数据排列于未点击广告数据之前的可能性，在重新投放广告数据的过程中，通常会选取排序靠前的部分广告数据进行投放，也就是说会优先将这些包含有未转化物品的广告数据投放至目标业务对象a1，因为目标业务对象a1存在兴趣的，那么相比于目标业务对象a1对其他不包含有未转化物品的广告数据产生转化行为的概率，目标业务对象a1对包 含有未转化物品的广告数据产生转化行为的概率较大，从而可以提高转化率。本申请实施例是通过每个业务对象的喜好，来进行精准的投放，故能够提升转化率。
需要说明的是，对于本申请实施例中所举例的广告标识(如2001)、对象标识(如a1)、物品标识(如FDY001)、是否存在点击行为的数值(1或0)、是否存在转化行为的数值(1或0)、记录时间戳(如14:00)以及浏览行为特征(如数据组[a1，2002，1，0，14:10])等各个参数，均是为便于理解所作出的举例说明，并不具备实际参考意义。
请参见图3，图3是本申请实施例提供的一种业务数据处理方法的流程示意图，其中，该业务数据处理方法可由计算机设备等电子设备执行，这里以该业务数据处理方法由计算机设备执行的情况为例进行说明。另外，这里的计算机设备可以是指业务服务器(如，上述图1中的业务服务器)，也可以是指终端设备(如，上述图1中终端设备集群中的任一终端设备)。如图3所示，该业务数据处理方法的流程可以至少包括步骤S101-步骤S103，下面对各步骤分别进行说明。
S101，获取N个已投放业务数据对应的投放日志；投放日志包括业务对象集合分别针对N个已投放业务数据的业务输出状态与业务转化状态；N为正整数；业务转化状态包括未转化状态；业务输出状态包括已输出状态。
在本申请实施例中，业务数据可以是指媒体数据(如广告数据)，投放可以是指曝光等推荐处理，而已投放业务数据可以是指已经向业务对象投放的媒体数据。其中，业务对象可以是指使用终端设备运行目标应用(如娱乐应用、社交应用、视频应用等)的业务用户在目标应用中的绑定账号，业务用户可以使用绑定账号登录目标应用，而目标应用也可以通过绑定账号判断业务用户是否登录、获取业务用户在目标应用中的相关行为数据等。应当理解，以业务数据为广告数据为例，上述如娱乐应用、社交应用、视频应用等目标应用，可作为广告投放平台，创建广告数据的广告主(也可称之为业务创建对象)可以将广告数据投放至广告投放平台中，在投放广告数据时，可针对性地选择业务用户群体进行投放，不同类型的广告数据可针对性地投放至目标应用中的不同业务用户群体，也就是说，会将不同类型的广告数据投放至目标应用中的不同的业务对象集合(称为业务用户群体对应的绑定账号)。
可以理解的是，将广告数据投放至业务用户群体后，这些广告数据可称之为已投放广告数据，这部分业务用户群体中的每个业务用户均可以播放观看已投放广告数据，每个业务用户在播放观看后还可以产生消费(如购买已投放广告数据中的商品)、下载等行为，而业务用户对已投放广告数据进行消费或下载的行为，可理解为转化行为，一个业务用户对一个已投放广告数据进行消费或下载，可以理解为是一个转化。以消费(如购买)行为为例，将一批广告数据曝光至业务用户群体A(可称之为业务用户集合A)后，针对某个广告数据，若该业务用户集合A中有50个用户通过该广告数据购买了某一产品，则该业务用户集合A针对该广告数据的转化数为50。
可以理解的是，计算机设备可通过业务用户群体的绑定账号(即业务对象)，获取已投放广告数据的相关数据(包括业务用户的播放观看行为、转化行为、观看停留时长、广告数据的名称、广告数据的类型和广告数据的曝光量等)。例如，可确定业务用户对某个已投放广告数据是否存在播放观看行为，而本申请实施例中的业务输出状态即可理解为业务用户是否对已投放广告数据存在播放观看行为(播放观看了即可理解为已投放广告数据输出至了该业务用户)，若某个业务用户对某个已投放广告数据存在播放观看行为，则该业务用户对应的业务对象针对该已投放广告数据的业务输出状态即可为已输出状态(也就是说，业务对象对应的业务用户观看了该广告数据)；若某个业务用户并未播放观看某个已投放广告数据，也就是该业务用户并未观看该已投放广告数据，那么该业务用户对应的业务对象针对该已投放广告数据的业务输出状态即可为未输出状态。
同理，计算机设备可通过业务用户群体(也可称为业务用户集合)的绑定账号，确定业务用户对某个已投放广告数据是否存在转化行为，而本申请实施例中的业务转化状态即可理解为业务用户是否对已投放广告数据存在转化行为。若某个业务用户针对某个已投放广告数据存在转化行为，则该业务用户对应的业务对象针对该已投放广告数据的业务转化状态即可为已转化状态；而若某个业务用户针对某个已投放广告数据不存在转化行为，则该业务用户对应的业务对象针对该已投放广告数据的业务转化状态即可为未转化状态。
可以理解的是，某个业务用户针对某个已投放广告数据的播放观看行为，可通过业务用户针对已投放广告数据在终端设备中的显示界面所执行的触发操作所产生；其中，该触发操作可以包含点击或者长按等接触性操作，也可以包含语音或手势等非接触性操作，这里将不对其进行限定。而在本申请实施例中，当某个业务用户对某个已投放广告数据产生播放观看行为时(如业务用户通过点击操作，播放了某个已投放广告数据)，计算机设备可以记录该业务用户产生点击操作的时间，当该业务用户发生了点击操作后，计算机设备会向业务用户对应的业务对象播放该已投放业务数据，那么该业务对象针对该已投放广告数据的业务输出状态也会由未输出状态变更为已输出状态，因此该点击操作也可以理解为是状态变更操作，该记录时间也可称为针对该状态变更操作的记录时间戳。
上述计算机设备所获取到的已投放广告数据的相关数据，可称为已投放广告数据的投放日志(也可称为广告日志)，投放日志中也可包括上述记录时间戳。
S102，根据业务对象集合分别针对N个已投放业务数据的业务输出状态与业务转化状态，建立目标业务对象针对未转化业务数据的未转化浏览行为特征；目标业务对象针对未转化业务数据的业务输出状态为已输出状态、且业务转化状态为未转化状态；N个已投放业务数据包括未转化业务数据，业务对象集合包括目标业务对象。
在本申请实施例中，投放日志中还可包括业务对象集合分别针对N个已投放业务数据的记录时间戳。以业务对象集合包括业务对象M i(i可为用于表征下标的任意数值，是一种索引变量，例如，i可为正整数、也可为分数等)、N个已投放业务数据包括已投放业务数据G b(b可为用于表征下标的任意数值，是一种索引变量，例如，b可为正整数、也可为分数等)为例，业务对象M i针对已投放业务数据G b的记录时间戳是指业务对象M i针对已投放业务数据G b的状态变更操作的操作记录时间戳，其中，状态变更操作可以用于指示业务对象M i针对已投放业务数据G b的业务输出状态由未输出状态转换为已输出状态。
计算机设备根据业务对象集合分别针对N个已投放业务数据的业务输出状态与业务转化状态，建立目标业务对象针对未转化业务数据的未转化浏览行为特征的的过程包括：计算机设备可获取业务对象M i的对象标识与已投放业务数据G b的业务标识；计算机设备可将业务对象M i的对象标识、已投放业务数据G b的业务标识、业务对象M i针对已投放业务数据G b的业务输出状态、业务对象M i针对已投放业务数据G b的业务转化状态、业务对象M i针对已投放业务数据G b的记录时间戳所组成的数据组，确定为业务对象M i针对已投放业务数据G b的浏览行为特征；随后，计算机设备可在业务对象集合针对N个已投放业务数据的浏览行为特征中，确定目标业务对象针对未转化业务数据的未转化浏览行为特征。
其中，计算机设备在业务对象集合针对N个已投放业务数据的浏览行为特征中，确定目标业务对象针对未转化业务数据的未转化浏览行为特征的过程包括：计算机设备可将业务对象集合针对N个已投放业务数据的浏览行为特征中，业务输出状态为已输出状态且业务转化状态为未转化状态的浏览行为特征，确定为未转化浏览行为特征；随后， 计算机设备可将未转化浏览行为特征中包含的对象标识所对应的业务对象，确定为目标业务对象，将未转化浏览行为特征中包含的业务标识所对应的已投放业务数据，确定为未转化业务数据。
应当理解，在获取到各个业务对象针对已投放业务数据的业务输出状态(如，是否存在播放观看行为)与业务转化状态(是否存在转化行为)后，可以根据业务对象的对象标识(用于表征业务对象，该对象标识可以为对象名称、对象ID等)、已投放业务数据的广告标识(用于表征业务数据，业务标识可以为业务名称、业务ID等)、业务输出状态、业务转化状态和产生播放观看行为的记录时间戳等数据，建立一个业务对象与一个已投放业务数据之间的数据组。该数据组即可为作为该业务对象针对该已投放业务数据的浏览行为特征。随后，可在建立的数据组中，提取出业务输出状态为已输出状态、且业务转化状态为未转化状态的数据组，而这些提取出来的数据组即可称为未转化浏览行为特征(或称为未转化数据组)，该未转化浏览行为特征中所包含的业务对象即可称为目标业务对象，该未转化浏览行为特征中所包含的已投放业务数据即可称为未转化业务数据。也就是说，若业务对象A播放观看了已投放业务数据A，但是未对已投放业务数据A产生转化行为，则该业务对象A可称为目标业务对象，该已投放业务数据A可称为未转化业务数据，该业务对象A与该已投放业务数据A所组成的数据组，可称为未转化浏览行为特征。
为便于理解，以下将举例阐述确定目标业务对象、未转化业务数据和未转化浏览行为特征的方式。以业务对象集合包括业务对象A、已投放业务数据包括已投放广告数据A为例进行说明；在一个时间段内，业务对象A在2021年8月7日19:00、2021年8月7日19:06均播放观看了该已投放广告数据A，该业务对象A针对该已投放广告数据A的业务输出状态为已输出状态，同时，该业务对象在2021年8月7日19:00、2021年8月7日19:06播放观看后，均未产生转化行为。如果以数值1表征已输出状态与已转化状态，以数值0表征未输出状态与未转化状态，那么根据该业务对象A的对象标识(即A)、已投放广告数据A的广告标识(即A)、业务输出状态(1或0)、业务转化状态(1或0)、记录时间戳(19:00、19:06)可组成2个数据组，包括数据组[A，A，1，0，19:00]、数据组[A，A，1，0，19:06]。
在本申请实施例中，在确定未转化浏览行为特征时，因为该数据组[A，A，1，0，19:00]与数据组A，A，1，0，19:06]中的业务输出状态为已输出状态、同时业务转化状态为未转化状态，从而计算机设备可将这两个数据组均确定为未转化浏览行为特征，可将该业务对象a确定为目标业务对象，将该已投放广告数据A确定为未转化业务数据。
S103，根据未转化浏览行为特征，确定目标业务对象与未转化物品共同对应的排序因子，根据排序因子，对包含有未转化物品的业务数据进行排序，得到序列业务数据，按照排序顺序从序列业务数据中选择目标业务数据，将目标业务数据投放至目标业务对象；未转化业务数据中包含未转化物品。
在本申请实施例中，业务数据可为广告数据，而每个广告数据用于推广某个商品，每个广告数据所推广的商品可称为业务物品。本申请实施例中的每个业务数据均可包含一个(可选的，也可为多个，当业务数据中包含多个业务物品时，可任选一个业务物品作为该业务数据中包含的业务物品)业务物品，而未转化业务数据中所包含的业务物品可称为未转化物品。根据未转化业务数据对应的未转化浏览行为特征，可确定目标业务对象与未转化物品共同对应的排序因子。其中，排序因子可以用于表征目标业务对象对于某个业务物品(如未转化物品)的感兴趣程度。
应当理解，一个未转化浏览行为特征中实际包含一个目标业务对象的对象标识与一个未转化业务数据的业务标识，而一个未转化业务数据中包含有一个未转化物品，多个 未转化业务数据中可能均会包含同一个未转化物品；而本申请实施例中计算机设备可以通过包含同一个未转化物品的多个(或一个)未转化业务数据，所分别对应的未转化浏览行为特征，确定出该目标业务对象与该未转化物品共同对应的排序因子。例如，如上述图2a-图2c所示，可以根据未转化浏览行为特征2000a、未转化浏览行为特征2000b、未转化浏览行为特征2000c和未转化浏览行为特征2000d，确定出目标业务对象a1与未转化物品“粉底液”共同对应的排序因子1。
也就是说，本申请实施例中，计算机设备可以确定出一个目标业务对象与一个未转化物品共同对应的排序因子。对于确定排序因子的实现方式，可以参见后续图4所对应的描述。
在本申请实施例中，计算机设备基于目标业务对象与某个未转化物品共同对应的排序因子，可对包含未转化物品的业务数据进行排序，并按照排序顺序选择目标业务数据，投放至目标业务对象。为便于理解，以包含有所述未转化物品的业务数据包括Q个目标未转化业务数据、Q个目标未转化业务数据中均包含同一未转化物品P x、且Q个目标未转化业务数据中包括目标未转化业务数据R s(表示Q个目标未转化业务数据中的每个目标未转化业务数据，s为任意用于表征下标的值，如s为取值范围是1至Q的正整数)为例进行说明。计算机设备对包含未转化物品的业务数据进行排序，得到序列业务数据的的过程包括：计算机设备可获取目标未转化业务数据R s针对业务对象集合的总投放量；根据总投放量，以及业务对象集合针对目标未转化业务数据R s的实时转化数量，可确定目标未转化业务数据R s对应的实时投放转化率；随后，计算机设备根据目标业务对象与未转化物品P x共同对应的排序因子，以及Q个目标未转化业务数据分别对应的实时投放转化率，可对包含有未转化物品P x的业务数据进行排序，得到序列业务数据。
应当理解，总投放量实际可理解为总曝光量(也可以理解为已投放业务数据的曝光次数，例如，某个业务数据向一千个业务对象进行曝光展示，该业务数据可称为已投放业务数据，该已投放业务数据的曝光量为一千次)；该实时转化数量可理解为业务数据在被投放后，实际的(真实的)发生转化行为的业务对象的数量，该实时转化数量也是衡量业务数据投放效果的指标，是指业务用户集合中，点击业务数据(如广告数据)并成为一个有效激活、注册或者付费用户的数量，通过实时转化数量也可得到实时转化比例(也可称为实时转化率，即该广告数据的实际转化次数(数量)除以广告数据的点击量(点击次数))。根据某个已投放业务数据的实时转化数量与总投放量，即可得到该已投放业务数据的实时投放转化率(如，将实时转化数量除以总投放量，即可得到实时投放转化率，也可称之为实时曝光转化率)。
在本申请实施例中，当包含有未转化物品的业务数据还包括包含未转化物品P x的待投放业务数据时，计算机设备根据目标业务对象与未转化物品P x共同对应的排序因子，以及Q个目标未转化业务数据分别对应的实时投放转化率，对包含有未转化物品P x的业务数据进行排序，得到序列业务数据的的过程包括：计算机设备可获取包含未转化物品P x的待投放业务数据，以及待投放业务数据对应的预估投放转化率；随后，计算机设备按照待排序已投放业务数据对应的实时投放转化率，与待投放业务数据对应的预估投放转化率之间的大小顺序，可对包含有未转化物品P x的业务数据进行排序，由此可得到初始序列业务数据；在初始序列业务数据中可按序获取预估投放转化率最大的K(K可为小于或等于W的正整数，W为待排序已投放业务数据与待投放业务数据的总数量)个业务数据；随后，可获取K个业务数据分别对应的期望转化单位投放资源、预估输出率以及预估转化率；根据目标业务对象与未转化物品P x共同对应的排序因子，以及K个业务数据分别对应的期望转化单位投放资源、预估输出率和预估转化率，可对 K个业务数据进行排序，得到序列业务数据。
需要说明的是，计算机设备可在N个已投放业务数据中，获取包含有未转化物品P x的待排序已投放业务数据；其中，包含有未转化物品P x的已投放业务数据(即待排序已投放业务数据)包括Q个目标未转化业务数据；另外，计算机设备可将待排序已投放业务数据与待投放业务数据确定为包含有未转化物品P x的业务数据。
这里，以K个业务数据包括业务数据R t(t为正整数)为例说明序列业务数据的获取过程，其中，K个业务数据包括业务数据R t是指业务数据R t为所述K个业务数据中的任一个。计算机设备根据目标业务对象与未转化物品P x共同对应的排序因子，以及K个业务数据分别对应的期望转化单位投放资源、预估输出率和预估转化率，对K个业务数据进行排序，得到序列业务数据的过程包括：计算机设备根据业务数据R t对应的期望转化单位投放资源、预估输出率和预估转化率，计算机设备可确定业务数据R t对应的初始投放单位消耗资源；将目标业务对象与未转化物品P x共同对应的排序因子，与业务数据R t对应的初始投放单位消耗资源进行运算处理，得到业务数据Rt对应的实时投放单位消耗资源；随后，计算机设备可根据K个实时投放单位消耗资源之间的大小顺序，对K个业务数据进行排序，由此可得到序列业务数据。
在本申请实施例中，计算机设备可从N个已投放业务数据中，获取到该目标业务对象并未产生播放观看行为(也就是并未产生点击行为，目标业务对象针对已投放业务数据的业务输出状态为未输出状态)的已投放业务数据，可将这些并未产生播放观看行为的已投放业务数据作为未输出业务数据(也可称为未播出业务数据，或者未点击业务数据)；也就是说，包含有未转化物品的业务数据还包括未输出业务数据。从而，计算机设备可将包含未转化物品P x的业务数据与该未输出业务数据一起进行排序，也就是说，N个已投放业务数据中还可包括未输出业务数据，那么计算机设备根据K个实时投放单位消耗资源之间的大小顺序，对K个业务数据进行排序，得到序列业务数据的过程包括：计算机设备获取未输出业务数据对应的期望转化单位投放资源、预估输出率、以及预估转化率；根据未输出业务数据对应的期望转化单位投放资源、预估输出率、以及预估转化率，确定未输出业务数据对应的初始投放单位消耗资源；按照未输出业务数据对应的初始投放单位消耗资源，与K个实时投放单位消耗资源之间的大小顺序，对K个业务数据与未输出业务数据进行排序，得到序列业务数据。
为便于理解本申请实施例中的包含有未转化物品P x的业务数据、以及计算机设备基于排序因子对包含有未转化物品P x的业务数据和未输出业务数据进行排序的过程，以下将结合上述图2a-图2c进行阐述。如图2a-图2c所示，未转化物品P x可为未转化物品“粉底液”，当未转化物品为“粉底液”时，确定出目标业务对象a1与未转化物品“粉底液”共同对应的排序因子1；同理，未转化物品P x可为未转化物品“纯牛奶”，当未转化物品为“纯牛奶”时，可确定出目标业务对象a1与未转化物品“纯牛奶”共同对应的排序因子2。在本申请实施例中，计算机设备可在N个已投放业务数据(即广告数据集合20)中，获取到包含未转化物品“粉底液”的已投放广告数据(包括已投放广告数据2001、已投放广告数据2002、已投放广告数据2003、已投放广告数据2004和已投放广告数据2008)，包含未转化物品“纯牛奶”的已投放广告数据(包括已投放广告数据2005和已投放广告数据2006)。那么所获取到的这些已投放广告数据均可称为包含未转化物品P x的已投放业务数据(也就是包含未转化物品P x的待排序已投放业务数据)。
在本申请实施例中，计算机设备也可以在候选业务数据集合(也就是由等待投放的候选业务数据所组成的集合，这些候选业务数据均未被投放至业务对象集合过，候选业务数据也可称为待投放业务数据，候选业务数据集合即候选广告数据集合。该候选业务数据集合中可包含未进行投放的历史业务数据，也可包含新创建的还未经过排序的也并 未进行投放的新增业务数据，历史业务数据与新增业务数据均可称为候选业务数据)中，获取到包含未转化物品P x的待投放业务数据(比如，包含未转化物品“粉底液”或未转化物品“纯牛奶”的待投放广告数据，可将这些包含未转化物品“粉底液”的待投放广告数据、以及上述包含未转化物品“粉底液”的待排序已投放业务数据，组成包含未转化物品“粉底液”的业务数据；可将这些包含未转化物品“纯牛奶”的待投放广告数据、以及包含未转化物品“纯牛奶”的待排序已投放业务数据，组成包含未转化物品“纯牛奶”的业务数据)。
在本申请实施例中，计算机设备可获取到已投放广告数据2001、已投放广告数据2002、已投放广告数据2003、已投放广告数据2004、已投放广告数据2008、已投放广告数据2005和已投放广告数据2006中分别对应的实时投放转化率；同时，因为包含未转化物品“粉底液”与未转化物品“纯牛奶”的待投放广告数据并未被投放，所以并不存在实时投放转化率，那么计算机设备可以获取到这些待投放广告数据中每个待投放广告数据的预估输出率(也可称为预估点击率(Predict Click Through Rate，PCTR)，预估点击率是指广告数据在某个情形下被投放后，在线广告***预估的该广告数据被点击的概率；也就是说，计算机设备将某个广告数据投放至某个业务对象集合后，预估的该业务对象集合将会点击该广告数据的业务对象数量与业务对象集合的总业务对象数量之间的比值)和预估转化率(Predict Conversion Rate，PCVR；预估转化率是指广告数据在某个情形下被点击后，在线广告***预估的被点击后的广告数据会发生转化的概率；也就是说，计算机设备将某个广告数据投放至某个业务对象集合后，在一部分业务对象对该广告数据进行点击后，这部分业务对象中，会对该广告数据发生转化行为的对象数量与这部分点击的业务对象的总对象数量的比值)，可根据每个待投放广告数据对应的预估点击率与预估转化率，确定出该待投放广告数据对应的预估投放转化率(比如，将预估点击率与预估转化率进行相乘运算处理来获得预估投放转化率)。
在本申请实施例中，计算机设备可以按照实时投放转化率与预估投放转化率之间的大小顺序，对这些包含未转化物品“粉底液”与未转化物品“纯牛奶”的业务数据进行排序，若包含未转化物品“粉底液”或未转化物品“纯牛奶”的待投放业务数据为待投放业务数据20010，则按照实时投放转化率与预估投放转化率由大到小的顺序进行排序后，所得到的初始序列业务数据为{已投放广告数据2008，已投放广告数据2001，已投放广告数据2006，已投放广告数据2003，已投放广告数据20010，已投放广告数据2002，已投放广告数据2004，已投放广告数据2005}；在本申请实施例中，计算机设备可按照业务场景需求，在该初始序列业务数据中获取位于前K个的已投放业务数据，以K为5为例，可提取得到前5个已投放广告数据：{已投放广告数据2008，已投放广告数据2001，已投放广告数据2006，已投放广告数据2003，已投放广告数据20010}。
在本申请实施例中，计算机设备可获取上述K个业务数据中每个业务数据的期望转化单位投放资源(可以是指广告主为广告数据竞标的价格，比如是指广告主对于一个转化对应的期望***格，也就是广告主决定的一个转化的费用)、预估输出率和预估转化率，并根据该期望转化单位投放资源、预估点击率、以及预估转化率，即可确定出各个业务数据分别对应的初始投放单位消耗资源。
示例性地，初始投放单位消耗资源可以是指将某个广告数据向一千个业务用户展示广告后，广告主需要支付的成本，初始投放单位消耗资源也可称为每千次曝光计费(Cost Per Mille，CPM)。在本申请实施例中，实际的CPM可以由广告主的期望转化单位投放资源、预估点击率、以及预估转化率所共同确定。为便于理解，如公式(1)所示，公式(1)是确定初始投放单位消耗资源的示例性方式。
CPM＝bid×PCTR×PCVR×1000     公式(1)；
其中，CPM可以是指某个广告数据实际的CPM，bid可以是指广告主(业务创建用户)对于该广告数据的期望转化单位投放资源，PCTR可以是指该广告数据的预估点击率，PCVR可以是指该广告数据的预估转化率。例如，某个广告数据R的预估点击率为0.1，预估转化率为0.1，期望转化单位投放资源为2，则该广告数据的CPM即可为20(2×0.1×0.1×1000)；也就是说，每一千次曝光需要向广告主收费20元。通常情况下，该实时的CPM即可作为广告数据的排序依据。
在本申请实施例中，计算机设备对包含有未转化物品的业务数据(广告数据)进行在排序时，在初始投放单位消耗资源的基础上，还需要加上排序因子的影响。例如，如果在如图2a-图2c的场景中，对于广告数据集合20，目标业务对象a1未产生播放观看行为的未输出业务数据为已投放广告数据2009，那么在进行排序时，计算机设备可以采用上述公式(1)，得到已投放广告数据2009的初始投放单位消耗资源；随后，计算机设备可在上述K个业务数据{已投放广告数据2008，已投放广告数据2001，已投放广告数据2006，已投放广告数据2003，待投放广告数据20010}中，确定出每个业务数据分别对应的初始投放单位消耗资源。在本申请实施例中，计算机可获取目标业务对象a1与未转化物品“粉底液”对应的排序因子1，以及目标业务对象a1与未转化物品“纯牛奶”对应的排序因子2；从而，计算机设备可在K个业务数据中获取包含未转化物品“粉底液”的广告数据，并将每个包含未转化物品“粉底液”的广告数据的初始投放单位消耗资源与排序因子1进行运算处理(例如，进行相加处理)，即可得到对应的实时投放单位消耗资源(即初始投放单位消耗资源与排序因子进行相加处理的结果)；同理，计算机设备通过排序因子2、以及包含未转化物品“纯牛奶”的广告数据对应的初始投放单位消耗资源，也可以确定包含未转化物品“纯牛奶”的广告数据对应的实时投放单位消耗资源。为便于理解，请参见公式(2)，公式(2)为确定实时投放单位消耗资源的方式。
CPM new＝bid×PCTR×PCVR×1000+quality i     公式(2)；
其中，CPM new可用于表征某个广告数据的实时投放单位消耗资源；bid×PCTR×PCVR×1000可用于表征该广告数据的初始投放单位消耗资源；quality i可用于表征该广告数据所包含的业务物品与某个目标业务对象共同对应的排序因子。通过公式(2)，可以得到K个业务数据{已投放广告数据2008，已投放广告数据2001，已投放广告数据2006，已投放广告数据2003，待投放广告数据20010}中各个广告数据分别对应的实时投放单位消耗资源，从而，计算机设备可按照该实时投放单位消耗资源与已投放广告数据2009的初始投放单位消耗资源之间的大小顺序，将K个业务数据与已投放广告数据2009(也可称为未输出广告数据2009)进行排序，得到序列业务数据(比如，序列广告数据)。而在排序得到序列业务数据后，可按照业务场景需求，在序列业务数据中按序选择目标业务数据(比如，目标广告数据)投放至目标业务对象a1。例如，计算机设备将上述K个业务数据{已投放广告数据2008，已投放广告数据2001，已投放广告数据2006，已投放广告数据2003，待投放广告数据20010}与已投放广告数据2009，按照K个实时投放单位消耗资源、以及已投放广告数据2009的初始投放单位消耗资源之间从大到小的顺序排序后，所得到的序列业务数据为{已投放广告数据2008，已投放广告数据2001，已投放广告数据2006，已投放广告数据2009，已投放广告数据2003，待投放广告数据20010}，那么此时，可以获取按照业务场景需求获取排序靠前的3个广告数据(即{已投放广告数据2008，已投放广告数据2001，已投放广告数据2006})，这3个广告数据{已投放广告数据2008，已投放广告数据2001，已投放广告数据2006}即可作为目标广告数据(目标业务数据)，可按序将这3个目标广告数据投放至目标业务对象a1。
应当理解，在对广告数据进行投放时，通常情况下，是按照预估点击率与预估转化 率对待投放广告数据进行排序，再按照业务场景需求获取排序靠前(预估点击率与预估转化率较大)的广告数据进行投放。然而在本申请实施例中，如若计算得到某个未转化物品的排序因子后，在对包含未转化物品的广告数据与未输出广告数据进行重新排序时，由于新增了额外的排序因子，那么对于包含未转化物品的广告数据，除了预估点击率与预估转化率的影响以外，还会存在排序因子的影响，排序因子可以提高包含未转化物品的广告数据排列于其他广告数据之前的概率，那么在向目标业务对象进行投放广告数据时，这些包含未转化物品的广告数据很可能会优先进行投放，从而能够提升投放的广告数据与目标业务对象的喜好的符号程度，进而能够提升目标业务对象产生转化行为的概率，由此可以提高转化率。
请参见图4，图4是本申请实施例提供的另一种业务数据处理方法的流程示意图。该流程可以对应于上述图3中的确定目标业务对象与未转化物品共同对应的排序因子的流程。如图4所示，该流程可以至少包括步骤401-步骤S403，下面对各步骤分别进行说明。
步骤401，对未转化业务数据进行业务识别，得到未转化业务数据与未转化物品之间的映射关系；映射关系用于指示未转化业务数据包含未转化物品。
应当理解，业务数据可为广告数据，计算机设备对未转化业务数据进行业务识别也就是对广告数据进行广告识别。其中，对广告数据进行广告识别可包括对广告数据进行语义内容识别，也可包括对广告数据进行图像内容识别。进行广告识别可为通过语音内容识别模型进行语义内容识别或通过图像内容识别模型进行图像内容识别，也可以对语音内容与图像内容均进行识别，然后选择准确率较高的模型进行广告识别；本申请实施例中的语义内容识别模型可为任一包括广告语义内容识别功能的模型，图像内容识别模型可为任一包括广告图像内容识别功能的模型，本申请实施例对此不做限定。计算机设备通过语义内容识别，可得到已投放广告数据与业务物品之间的映射关系：R i——>C i，其中，R i可用于表征某个已投放广告数据的广告标识，C i可用于表征某个业务物品的物品标识，该映射关系可表征该已投放广告数据中包含该业务物品C i。应当理解，未转化业务数据中包含的业务物品可称为未转化物品。
步骤402，获取未转化物品的物品标识，将未转化浏览行为特征中包含的未转化业务数据的业务标识，替换为未转化物品的物品标识，将替换后的未转化浏览行为特征，确定为目标业务对象针对未转化物品的目标浏览行为特征。
在本申请实施例中，计算机设备可将未转化浏览行为特征中未转化业务数据的业务标识替换为未转化物品的物品标识，由此可得到目标业务对象针对未转化物品的目标浏览行为特征。例如，如上述图2a-图2c所对应的场景所示，以未转化浏览行为特征为[a1，2001，1，0，14:00]和[a1，2002，1，0，14:10]为例，该未转化浏览行为特征中广告数据2001与广告数据2002均包含未转化物品“粉底液”，该未转化物品“粉底液”的物品标识为FDY001，进行标识替换后，可得到目标业务对象a1针对未转化物品“粉底液”的目标浏览行为特征为[a1，FDY001，1，0，14:00]和[a1，FDY001，1，0，14:10]。
步骤403，根据目标浏览行为特征，确定目标业务对象与未转化物品共同对应的排序因子。
在本申请实施例中，计算机设备根据目标业务对象针对未转化物品的目标浏览行为特征，即可确定目标业务对象与未转化物品共同对应的排序因子。这里，包含未转化物品的未转化业务数据包括Q(Q可为小于或等于N的整数)个目标未转化业务数据；Q个目标未转化业务数据中均包含同一未转化物品P x(x为任意用于表征下标的值，x可为整数、分数、字母等)；目标业务对象针对未转化物品的目标浏览行为特征包括目标业务对象针对未转化物品P x的Q个目标浏览行为特征，排序因子的获取过程包括：计 算机设备可统计Q个目标浏览行为特征的特征总数量，可将特征总数量确定为目标业务对象针对未转化物品P x的总输出次数；从Q个目标浏览行为特征分别包含的记录时间戳中，获取最早记录时间戳；根据当前时间戳、总输出次数、以及最早记录时间戳，即可确定目标业务对象与未转化物品P x共同对应的排序因子。
为便于理解确定排序因子的方法，以下将结合上述图2a-图2c所对应的描述进行说明。如图2a和图2b所示，对于目标业务对象a1而言，未转化广告数据2001、未转化广告数据2002、未转化广告数据2003、未转化广告数据2004、未转化广告数据2005、未转化广告数据2006均为目标业务对象a1的未转化广告数据，而在这些未转化广告数据中，未转化广告数据2001、未转化广告数据2002、未转化广告数据2003和未转化广告数据2004均包含相同的未转化物品“粉底液”，则上述Q个目标未转化业务数据即可为未转化广告数据2001、未转化广告数据2002、未转化广告数据2003和未转化广告数据2004；上述未转化物品P x即可为该未转化物品“粉底液”；上述目标业务对象针对未转化物品P x的Q个目标浏览行为特征也就是目标业务对象a1针对未转化物品“粉底液”的目标浏览行为特征2000a’(通过未转化浏览行为特征2000a得到)、目标业务对象a1针对未转化物品“粉底液”的目标浏览行为特征2000b’(通过未转化浏览行为特征2000b得到)、目标业务对象a1针对未转化物品“粉底液”的目标浏览行为特征2000c’(通过未转化浏览行为特征2000c得到)、以及目标业务对象a1针对未转化物品“粉底液”的目标浏览行为特征2000d’(通过未转化浏览行为特征2000d得到)。
同理，对于目标业务对象a1而言，在这些未转化广告数据中，未转化广告数据2005和未转化广告数据2006均包含相同的未转化物品“纯牛奶”，则上述Q个目标未转化业务数据即可为未转化广告数据2005和未转化广告数据2006；上述未转化物品P x即可为该未转化物品“纯牛奶”；上述目标业务对象针对未转化物品P x的Q个目标浏览行为特征也就是目标业务对象a1针对未转化物品“纯牛奶”的目标浏览行为特征2000e’(通过未转化浏览行为特征2000e得到)、以及目标业务对象a1针对未转化物品“纯牛奶”的目标浏览行为特征2000f’(通过未转化浏览行为特征2000f得到)。
在本申请实施例中，计算机设备根据目标业务对象针对某个未转化物品的目标浏览行为特征，可以确定出该目标业务对象与这个未转化物品共同对应的排序因子。例如，如图2c所示，计算机设备根据目标业务对象a1针对未转化物品“粉底液”的目标浏览行为特征(包括目标浏览行为特征2000a’、目标浏览行为特征2000b’、目标浏览行为特征2000c’和目标浏览行为特征2000d’)，可以确定出目标业务对象a1与未转化物品“粉底液”共同对应的排序因子1，包括：计算机设备根据目标业务对象a1针对未转化物品“粉底液”的目标浏览行为特征，确定出目标业务对象a1针对该未转化物品的总点击次数(称为总输出次数)；其中，因为一个目标浏览行为特征对应的是一次点击行为(也就是目标业务对象针对某个未转化广告数据产生了播放观看行为)，故该总点击次数实际也就相当于是这些目标浏览行为特征的特征总数量。而因为在目标业务对象a1针对未转化物品“粉底液”的每个目标浏览行为特征中，均包含有业务输出状态的记录时间戳(即产生播放观看行为的时间或时刻)，从而计算机设备可在目标业务对象a1针对未转化物品“粉底液”的目标浏览行为特征中，获取到最小记录时间戳(也可称为最早记录时间戳或最早时刻，该最小记录时间戳即最早点击时间(或称为最早点击时刻)，也即最早产生播放观看行为的时间)；接着，计算机设备根据当前时间(当前所处的时刻，称为当前时间戳)、最小记录时间戳、以及总点击次数，即可确定出目标业务对象a1与未转化物品“粉底液”共同对应的排序因子。
为便于理解，请参见公式(3)，公式(3)是根据当前时间、最小记录时间戳和总点击次数确定排序因子的方式，如下所示。
Figure PCTCN2022099533-appb-000001
其中，公式(3)中的∑click i可用于表征某个目标业务对象针对某个业务物品的总点击次数(如上述目标业务对象针对未转化物品“粉底液”的总点击次数)；公式(3)中的t可用于表征当前时间(或称为当前时刻)；公式(3)中的t 0可用于表征目标业务对象针对某个业务物品的最早发生点击行为的时刻；公式(3)中的e α是一个指数函数，quality i可用于表征目标业务对象针对某个业务物品的排序因子。
在本申请实施例中，通过业务对象集合分别针对N个已投放业务数据的业务输出状态和业务转化状态，建立目标业务对象针对未转化业务数据的未转化浏览行为特征，并根据该未转化浏览行为特征确定出目标业务对象与未转化物品(即未转化业务数据中包括的业务物品)共同对应的排序因子；随后，根据该排序因子对包含有未转化物品的业务数据进行排序，并按序选择目标业务数据投放至目标业务对象。可以理解的是，上述向目标业务对象投放目标业务数据的过程中，目标业务对象对未转化业务数据并未发生转化行为，但因为未转化业务数据的业务输出状态已输出状态，所以仍然可表明该目标业务对象是对该未转化业务数据中的业务物品存在兴趣；也就是说，本申请实施例可以根据目标业务对象对已投放业务数据的实时反馈数据，来确定出对于目标业务对象而言，为已输出状态且为未转化状态的未转化业务数据，而这些未转化业务数据可表明目标业务对象的潜在喜好，按照喜好将包含未转化物品的业务数据进行排序并投放，使得投放的业务数据是符合目标业务对象的喜好的，所以能够提升精准度，可以减少数据传输成本，而且因为投放的数据是符合目标业务对象的喜好的，能够提高目标业务对象产生转化行为的概率，从而可以提高的转化率。综上，本申请实施例可以提高业务数据的投放精准度，节约数据传输成本，提高业务数据的转化率。
可以理解的是，通过上述图3和图4所对应的描述可知，针对某个目标业务对象，可将序列业务数据中的目标业务数据(目标业务数据可为完整的序列业务数据，也可为序列业务数据中的部分业务数据)在目标时间段内投放至该目标业务对象；在这个投放的过程中，计算机设备可以获取到该目标业务对象针对该目标业务数据中每个业务数据的浏览反馈数据，从而计算机设备能够基于该浏览反馈数据进行后续处理(例如，停止序列业务数据的投放；或者进行相似业务数据推荐处理等)。为便于理解，请参见图5，图5是本申请实施例提供的一种基于目标业务对象的浏览反馈数据，进行相似业务数据推荐处理的流程示意图。如图5所示，该流程可以至少包括以下步骤S501和步骤S502，下面对各步骤分别进行说明。
S501，在目标时间段内，获取目标业务对象针对目标业务数据的浏览反馈数据。
在本申请实施例中，针对目标业务数据中的每一个业务数据(广告数据)，目标业务对象对应的业务用户可通过点击触发操作，进行播放观看，并在播放观看后产生转化行为，那么此时该浏览反馈数据即可为转化反馈行为数据；相应地，目标业务对象对应的业务用户也可以通过点击负向反馈控件(如关闭控件、停止控件、不感兴趣控件和广告吐槽控件等)，来表征自己对某个业务数据的负向喜好，那么此时该浏览反馈数据即可为负向反馈行为数据；相应地，目标业务对象对应的业务用户在播放观看某个业务数据后，那么也就是对该业务数据进行了浏览观看，计算机设备也可以统计该业务用户在目标时间段内针对业务数据的目标浏览时长，那么也就是说，该浏览反馈数据可为目标业务对象针对某个业务数据的目标浏览时长。从而，浏览反馈数据包括以下至少一项：目标业务对象针对目标业务数据的转化反馈行为数据，目标业务对象针对目标业务数据的负向反馈行为数据，目标业务对象针对目标业务数据的目标浏览时长。
S502，根据浏览反馈数据，对目标业务对象进行相似业务数据投放处理。
在本申请实施例中，当浏览反馈数据包括目标业务对象针对序列业务数据的转化反 馈行为数据时，计算机设备对目标业务对象进行相似业务数据推荐处理的过程包括：计算机设备从目标业务数据中，获取转化反馈行为数据对应的转化业务数据；并获取转化业务数据中所包括的转化物品，获取转化物品的第一相似物品；以及将第一相似物品所属的业务数据作为目标相似业务数据，将目标相似业务数据投放至目标业务对象。应当理解，在将目标业务数据投放后，若目标业务对象对包含业务物品“美白面膜”的业务数据A产生了转化行为，那么该业务物品“美白面膜”可称为转化物品，此时可表明该目标业务对象对业务物品“美白面膜”是存在兴趣的，那么此时计算机设备可按照目标业务对象的喜好，向目标业务对象投放包含相似物品(如同类型的物品，如美白精华和美白霜等)对应的业务数据。
在本申请实施例中，当浏览反馈数据包括目标业务对象针对目标业务数据的负向反馈行为数据时，计算机设备对目标业务对象进行相似业务数据推荐处理的过程包括：计算机设备从目标业务数据中，获取负向反馈行为数据对应的负向反馈业务数据；并获取负向反馈业务数据中所包括的负向反馈物品，获取负向反馈物品的第二相似物品；以及将第二相似物品所属的业务数据作为目标相似业务数据，在针对目标业务对象的业务数据投放过程中，对目标相似业务数据进行过滤。应当理解，在将目标业务数据投放后，若目标业务对象对包含业务物品“美白面膜”的业务数据A产生了负向反馈行为数据，那么该业务物品“美白面膜”可称为负向反馈物品，此时可表明该目标业务对象对物品“美白面膜”是不存在兴趣的，那么此时计算机设备可按照目标业务对象的喜好，在向目标业务对象进行新一轮投放(在目标时间段的下一个时间段投放)时，不再向目标业务对象投放包含相似物品(如同类型的物品，如美白精华、美白霜等)对应的业务数据。
在本申请实施例中，当浏览反馈数据包括目标业务对象针对目标业务数据的浏览时长时，计算机设备对目标业务对象进行相似业务数据推荐处理的过程包括：计算机设备获取目标业务对象针对目标业务数据的历史平均浏览时长；当目标浏览时长大于历史平均浏览时长时，将目标业务数据中包括的业务物品确定为潜在转化物品，并将潜在转化物品的第三相似物品所属的业务数据作为目标相似业务数据，以及将目标相似业务数据投放至目标业务对象；其中，计算机设备在将目标业务数据投放至目标业务对象的过程中，能够获取到目标业务对象针对目标业务数据的目标浏览时长。应当理解，若目标业务对象对目标业务数据中某个业务数据的目标浏览时长，大于该目标业务对象针对该业务数据的历史平均浏览时长，则可表明在投放过程中，该目标业务对象对该业务数据的兴趣度增大，在新一轮的投放过程中，可以向目标业务对象投放包含同类型的相似物品的业务数据。
应当理解，在本申请实施例中，可根据目标业务对象针对序列业务数据的实时反馈数据(如浏览反馈数据)，并基于该浏览反馈数据确定目标业务对象的喜好进行相应业务处理。在目标业务对象发生了转化行为时，可以向目标业务对象继续投放同类型的业务数据；在目标业务对象发生了负向反馈行为时，可以及时停止投放序列业务数据，并在一定时间内不再像目标业务对象投放同类型的业务数据；在目标业务对象的浏览时长有所增长时，也可以继续向目标业务对象投放同类型的业务数据。也就是说，通过目标业务对象的实时反馈数据，能够提升向目标业务对象投放业务数据的精准度，提高目标业务对象对应的转化率，提升用户体验感。
请参见图6，图6是本申请实施例提供的一种***结构图。如图6所示，该***结构中可包括广告内容识别模块、特征确定模块、同类广告确定模块、点击率、转化率预估模块、排序因子计算模块、广告投放模块、反馈数据收集模块、行为时间过滤模块和负向过滤模块。
其中，广告内容识别模块可用于对广告数据进行语义内容识别和图像内容识别中的 至少一种识别。特征确定模块可用于确定目标业务对象(实际可以指发生点击行为但未产生转化行为的业务用户对应的业务对象，也可以称为点击未转化用户对应的业务对象)针对某个未转化物品的浏览行为特征。同类广告确定模块可用于获取包含同一个未转化物品的所有广告数据。点击率、预估量预估模块可用于对各个广告数据的点击率与转化率进行预估。排序因子计算模块可用于确定某个目标业务对象与某个未转化物品共同对应的排序因子。广告投放模块可基于排序因子计算模块所计算得到的排序因子，对广告数据进行排序并按业务场景需求获取部分广告数据进行投放。反馈数据收集模块可用于收集各个业务对象针对各个业务数据的浏览反馈数据(例如，点击行为数据和转化行为数据等)。应当理解，本申请实施例可基于业务对象的浏览反馈数据，确定出点击未转化用户和点击未转化用户对应的未转化广告数据等，而因为业务对象的转化需求是存在有效期的，因此在本申请实施例中可以提取用户点击未转化行为的时间窗口；比如，采取最近T天内的点击未转化行为，即假设当前时刻记为now，则可提取在now-T至now的时间段内针对某个已投放广告数据发生点击未转化行为的业务对象，该业务对象即可作为目标业务对象，该已投放广告数据即可作为未转化业务数据。负向过滤模块可用于将业务对象产生了负向反馈行为数据的广告数据进行过滤。
图6中的各个模块可用于实现上述图3-图5所对应的流程，对于各个模块的实现方式，可以参见上述图3-图5所对应的描述，这里将不再进行赘述。
请参见图7，图7是本申请实施例提供的一种业务数据处理装置的结构示意图。该业务数据处理装置可以是运行于计算机设备等电子设备中的一个计算机程序(包括程序代码)，例如该业务数据处理装置为一个应用软件；该业务数据处理装置可以用于执行图3所示的方法。如图7所示，该业务数据处理装置71可以包括：日志获取模块11、特征建立模块12、排序因子确定模块13、数据排序模块14、以及数据投放模块15。
日志获取模块11，配置为获取N个已投放业务数据对应的投放日志；所述投放日志包括业务对象集合分别针对所述N个已投放业务数据的业务输出状态与业务转化状态；N为正整数；所述业务转化状态包括未转化状态；所述业务输出状态包括已输出状态；
特征建立模块12，配置为根据所述业务对象集合分别针对所述N个已投放业务数据的所述业务输出状态与所述业务转化状态，建立目标业务对象针对未转化业务数据的未转化浏览行为特征；所述目标业务对象针对所述未转化业务数据的所述业务输出状态为所述已输出状态、且所述业务转化状态为所述未转化状态；所述N个已投放业务数据包括所述未转化业务数据，所述业务对象集合包括所述目标业务对象；
排序因子确定模块13，配置为根据所述未转化浏览行为特征，确定所述目标业务对象与未转化物品共同对应的排序因子；未转化业务数据中包含未转化物品。
数据排序模块14，配置为根据所述排序因子，对包含有所述未转化物品的业务数据进行排序，得到序列业务数据；
数据投放模块15，配置为按照排序顺序从所述序列业务数据中选择目标业务数据，将所述目标业务数据投放至所述目标业务对象。
其中，日志获取模块11、特征建立模块12、排序因子确定模块13、数据排序模块14、以及数据投放模块15的实现方式，可以参见上述图3所对应实施例中步骤S101-步骤S103的描述，这里将不再进行赘述。
在本申请实施例中，所述投放日志还包括所述业务对象集合分别针对所述N个已投放业务数据的记录时间戳；所述业务对象集合包括业务对象M i，所述N个已投放业务数据包括已投放业务数据G b，所述业务对象M i针对所述已投放业务数据G b的所述记录时间戳，是指所述业务对象M i针对所述已投放业务数据G b的状态变更操作的操作记录 时间戳，所述状态变更操作用于指示所述业务对象M i针对所述已投放业务数据G b的所述业务输出状态由未输出状态转换为所述已输出状态；i、b均为正整数；
特征建立模块12，包括：数据获取单元121与特征确定单元122。
数据获取单元121，配置为获取所述业务对象M i的对象标识与所述已投放业务数据G b的业务标识；
特征确定单元122，配置为将所述业务对象M i的所述对象标识、所述已投放业务数据G b的所述业务标识、所述业务对象M i针对所述已投放业务数据G b的所述业务输出状态、所述业务对象M i针对所述已投放业务数据G b的所述业务转化状态、以及所述业务对象Mi针对所述已投放业务数据G b的所述记录时间戳所组成的数据组，确定为所述业务对象M i针对所述已投放业务数据G b的浏览行为特征；
特征确定单元122，还配置为在所述业务对象集合针对所述N个已投放业务数据的所述浏览行为特征中，确定所述目标业务对象针对所述未转化业务数据的所述未转化浏览行为特征。
其中，数据获取单元121与特征确定单元122的实现方式，可以参见上述图3所对应的步骤S102的描述，这里将不再进行赘述。
在本申请实施例中，特征确定单元122，还配置为在所述业务对象集合针对所述N个已投放业务数据的所述浏览行为特征中，将所述业务输出状态为所述已输出状态、且所述业务转化状态为所述未转化状态的所述浏览行为特征，确定为所述未转化浏览行为特征；
特征确定单元122，还配置为将所述未转化浏览行为特征中包含的所述对象标识所对应的业务对象，确定为所述目标业务对象，在包含所述目标业务对象的所述对象标识的所述未转化浏览行为特征中，将包含的所述业务标识所对应的已投放业务数据，确定为所述目标业务对象针对的所述未转化业务数据，以得到所述目标业务对象针对所述未转化业务数据的所述未转化浏览行为特征。
在本申请实施例中，排序因子确定模块13可以包括：数据识别单元131、标识替换单元132、以及排序因子确定单元133。
数据识别单元131，配置为对未转化业务数据进行业务识别，得到未转化业务数据与未转化物品之间的映射关系；映射关系用于指示未转化业务数据包含未转化物品；
标识替换单元132，配置为获取所述未转化物品的物品标识，将所述未转化浏览行为特征中包含的所述未转化业务数据的业务标识，替换为所述未转化物品的所述物品标识，将替换后的未转化浏览行为特征，确定为所述目标业务对象针对所述未转化物品的目标浏览行为特征；
排序因子确定单元133，配置为根据所述目标浏览行为特征，确定所述目标业务对象与所述未转化物品共同对应的所述排序因子。
其中，数据识别单元131、标识替换单元132、以及排序因子确定单元133的实现方式，可以参见上述图4所对应的步骤S401-步骤S403的描述，这里将不再进行赘述。
在本申请实施例中，排序因子确定单元133，还配置为统计所述Q个目标浏览行为特征的特征总数量，将所述特征总数量确定为所述目标业务对象针对所述未转化物品P x的总输出次数；
排序因子确定单元133，还配置为从所述Q个目标浏览行为特征分别包含的记录时间戳中，获取最早记录时间戳；
排序因子确定单元133，还配置为根据当前时间戳、所述总输出次数、以及所述最早记录时间戳，确定所述目标业务对象与所述未转化物品P x共同对应的所述排序因子。
在本申请实施例中，包含有所述未转化物品的业务数据包括Q个目标未转化业务数 据；所述Q个目标未转化业务数据中均包含同一未转化物品Px；Q为小于或等于N的正整数，x为正整数；
数据排序模块14可以包括：投放转化率确定单元141与排序单元142。
投放转化率确定单元141，配置为获取目标未转化业务数据R s针对所述业务对象集合的总投放量，所述目标未转化业务数据R s为所述Q个目标未转化业务数据中的任一个；
投放转化率确定单元141，还配置为根据所述总投放量、以及所述业务对象集合针对所述目标未转化业务数据R s的实时转化数量，确定所述目标未转化业务数据R s对应的实时投放转化；
排序单元142，配置为根据所述排序因子、以及所述Q个目标未转化业务数据分别对应的所述实时投放转化率，对包含有所述未转化物品P x的业务数据进行排序，得到所述序列业务数据。
其中，投放转化率确定单元141与排序单元142的实现方式，可以参见上述图3所对应的步骤S103的描述，这里将不再进行赘述。
在本申请实施例中，排序单元142，还配置为获取所述待投放业务数据对应的预估投放转化率；包含有所述未转化物品的业务数据还包括包含所述未转化物品P x的待投放业务数据；
排序单元142，还配置为按照所述Q个目标未转化业务数据分别对应的所述实时投放转化率，与所述待投放业务数据对应的所述预估投放转化率之间的大小顺序，对所述包含有所述未转化物品P x的业务数据进行排序，得到初始序列业务数据；
排序单元142，还配置为从所述初始序列业务数据中，按序获取所述预估投放转化率最大的K个业务数据；K为小于W的正整数；W为所述Q个目标未转化业务数据与包含所述未转化物品P x的所述待投放业务数据的总数量；
排序单元142，还配置为获取所述K个业务数据分别对应的期望转化单位投放资源、预估输出率、以及预估转化率；
排序单元142，还配置为根据所述排序因子，以及所述K个业务数据分别对应的所述期望转化单位投放资源、所述预估输出率和所述预估转化率，对所述K个业务数据进行排序，得到所述序列业务数据。
在本申请实施例中，业务数据R t为所述K个业务数据中的任一个，t为正整数；
排序单元142，还配置为根据所述业务数据R t对应的所述期望转化单位投放资源、所述预估输出率、以及所述预估转化率，确定所述业务数据R t对应的初始投放单位消耗资源；
排序单元142，还配置为将所述排序因子与所述目标业务数据R t对应的所述初始投放单位消耗资源进行运算处理，得到所述目标业务数据R t对应的实时投放单位消耗资源；
排序单元142，还配置为根据K个实时投放单位消耗资源之间的大小顺序，对所述K个业务数据进行排序，得到所述序列业务数据。
在本申请实施例中，包含有所述未转化物品的业务数据还包括未输出业务数据；目标业务对象针对未输出业务数据的业务输出状态为未输出状态；
排序单元142，还配置为获取所述未输出业务数据对应的所述期望转化单位投放资源、所述预估输出率、以及所述预估转化率；
排序单元142，还配置为根据所述未输出业务数据对应的所述期望转化单位投放资源、所述预估输出率、以及所述预估转化率，确定所述未输出业务数据对应的所述初始投放单位消耗资源；
排序单元142，还配置为按照未输出业务数据对应的初始投放单位消耗资源，与K 个实时投放单位消耗资源之间的大小顺序，对K个业务数据与未输出业务数据进行排序，得到序列业务数据。
在本申请实施例中，该业务数据处理装置71还可以包括：反馈数据获取模块16与数据推荐模块17。
反馈数据获取模块16，配置为在目标时间段内，获取目标业务对象针对目标业务数据的浏览反馈数据；
数据推荐模块17，配置为根据浏览反馈数据，对目标业务对象进行相似业务数据投放处理。
其中，反馈数据获取模块16与数据推荐模块17的实现方式，可以参见上述图5所对应的步骤S501-步骤S502的描述，这里将不再进行赘述。
在本申请实施例中，浏览反馈数据包括目标业务对象针对目标业务数据的转化反馈行为数据；
数据推荐模块17可以包括：转化物品获取单元171与第一数据推荐单元172。
转化物品获取单元171，配置为从所述目标业务数据中，获取所述转化反馈行为数据对应的转化业务数据；
转化物品获取单元171，还配置为获取转化业务数据中所包括的转化物品，获取转化物品的第一相似物品；
第一数据推荐单元172，配置为将第一相似物品所属的业务数据作为目标相似业务数据，将目标相似业务数据投放至目标业务对象。
其中，对于转化物品获取单元171与第一数据推荐单元172的实现方式，可以参见上述图5所对应的步骤S502中的描述，这里将不再进行赘述。
在本申请实施例中，浏览反馈数据包括目标业务对象针对目标业务数据的负向反馈行为数据；
数据推荐模块17可以包括：负向反馈物品获取单元173与第二数据推荐单元174。
负向反馈物品获取单元173，配置为从所述目标业务数据中，获取所述负向反馈行为数据对应的负向反馈业务数据；
负向反馈物品获取单元173，还配置为获取负向反馈业务数据中所包括的负向反馈物品，获取负向反馈物品的第二相似物品；
第二数据推荐单元174，配置为将第二相似物品所属的业务数据作为目标相似业务数据，在针对目标业务对象的业务数据投放过程中，对目标相似业务数据进行过滤。
其中，负向反馈物品获取单元173与第二数据推荐单元174的实现方式，可以参见上述图5所对应的步骤S502中的描述，这里将不再进行赘述。
在本申请实施例中，浏览反馈数据包括目标业务对象针对目标业务数据的浏览时长；
数据推荐模块17可以包括：时长获取单元175与第三数据推荐单元176。
时长获取单元175，配置为获取目标业务对象针对目标业务数据的历史平均浏览时长；
第三数据推荐单元176，配置为当目标浏览时长大于历史平均浏览时长时，将目标业务数据中包括的业务物品确定为潜在转化物品，将潜在转化物品的第三相似物品所属的业务数据作为目标相似业务数据，将目标相似业务数据投放至目标业务对象。
其中，时长获取单元175与第三数据推荐单元176的实现方式，可以参见上述图5所对应的步骤S502中的描述，这里将不再进行赘述。
请参见图8，图8是本申请实施例提供的一种计算机设备的结构示意图。如图8所示，上述图7所对应实施例中的业务数据处理装置71可以应用于上述计算机设备8000(称为电子设备)，上述计算机设备8000可以包括：处理器8001、网络接口8004和存 储器8005，此外，上述计算机设备8000还包括：用户接口8003和至少一个通信总线8002。其中，通信总线8002用于实现这些组件之间的连接通信。其中，用户接口8003可以包括显示屏(Display)和键盘(Keyboard)，用户接口8003还可以包括标准的有线接口、无线接口。网络接口8004可以包括标准的有线接口和无线接口(如WI-FI接口)。存储器8005可以是高速RAM存储器，也可以是非不稳定的存储器(Non-Volatile Memory)，例如至少一个磁盘存储器。存储器8005可选的还可以是至少一个位于远离前述处理器8001的存储装置。如图8所示，作为一种计算机可读存储介质的存储器8005中可以包括操作***、网络通信模块、用户接口模块、以及设备控制应用程序。
在图8所示的计算机设备8000中，网络接口8004可提供网络通讯功能；而用户接口8003主要用于为用户提供输入的接口；而处理器8001可以用于调用存储器8005中存储的设备控制应用程序，以实现本申请实施例提供的业务数据处理方法。
本申请实施例中所描述的计算机设备8000可执行图3到图5所对应的对该业务数据处理方法的描述，也可执行图7所对应的对该业务数据处理装置71的描述，在此不再赘述。另外，对采用相同方法的有益效果描述，也不再进行赘述。
本申请实施例还提供了一种计算机可读存储介质，且上述计算机可读存储介质中存储有计算机设备8000所执行的计算机程序，且上述计算机程序包括程序指令，当上述处理器执行上述程序指令时，能够执行图3到图5所对应的业务数据处理方法，因此，这里将不再进行赘述。另外，对采用相同方法的有益效果描述，也不再进行赘述。
上述计算机可读存储介质可以是本申请实施例提供的业务数据处理装置或者计算机设备的内部存储单元，例如计算机设备的硬盘或内存。该计算机可读存储介质也可以是该计算机设备的外部存储设备，例如该计算机设备上配备的插接式硬盘，智能存储卡(Smart Media Card，SMC)，安全数字(Secure Digital，SD)卡，闪存卡(Flash Card)等。在本申请实施例中，该计算机可读存储介质还可以既包括该计算机设备的内部存储单元也包括外部存储设备。该计算机可读存储介质用于存储该计算机程序以及该计算机设备所需的其他程序和数据。该计算机可读存储介质还可以用于暂时地存储已经输出或者将要输出的数据。
本申请实施例提供了一种计算机程序产品或计算机程序，该计算机程序产品或计算机程序包括计算机指令，该计算机指令存储在计算机可读存储介质中。计算机设备的处理器从计算机可读存储介质读取该计算机指令，处理器执行该计算机指令，使得该计算机设备执行本申请实施例提供的业务数据处理方法。
本申请实施例的说明书和权利要求书及附图中的术语“第一”、“第二”和“第三”等是用于区别不同对象，而非用于描述特定顺序。此外，术语“包括”以及它们任何变形，意图在于覆盖不排他的包含。例如包含了一系列步骤或单元的过程、方法、装置、产品或设备没有限定于已列出的步骤或模块，而是可选地还包括没有列出的步骤或模块，或可选地还包括对于这些过程、方法、装置、产品或设备固有的其他步骤单元。
本领域普通技术人员可以意识到，结合本申请实施例中所公开的描述的各示例的单元及算法步骤，能够以电子硬件、计算机软件或者二者的结合来实现，为了清楚地说明硬件和软件的可互换性，在上述说明中已经按照功能一般性地描述了各示例的组成及步骤。这些功能究竟以硬件还是软件方式来执行，取决于技术方案的特定应用和设计约束条件。专业技术人员可以对每个特定的应用来使用不同方法来实现所描述的功能，但是这种实现不应认为超出本申请的范围。
本申请实施例提供的方法及相关装置是参照本申请实施例提供的方法流程图和/或结构示意图来描述的，可由计算机程序指令实现方法流程图和/或结构示意图的每一流程和/或方框、以及流程图和/或方框图中的流程和/或方框的结合。这些计算机程序指令可 提供到通用计算机、专用计算机、嵌入式处理机或其他可编程数据处理设备的处理器以产生一个机器，使得通过计算机或其他可编程数据处理设备的处理器执行的指令产生用于实现在流程图一个流程或多个流程和/或结构示意图一个方框或多个方框中指定的功能的装置。这些计算机程序指令也可存储在能引导计算机或其他可编程数据处理设备以特定方式工作的计算机可读存储器中，使得存储在该计算机可读存储器中的指令产生包括指令装置的制造品，该指令装置实现在流程图一个流程或多个流程和/或结构示意图一个方框或多个方框中指定的功能。这些计算机程序指令也可装载到计算机或其他可编程数据处理设备上，使得在计算机或其他可编程设备上执行一系列操作步骤以产生计算机实现的处理，从而在计算机或其他可编程设备上执行的指令提供用于实现在流程图一个流程或多个流程和/或结构示意一个方框或多个方框中指定的功能的步骤。
可以理解的是，在本申请实施例中，涉及到投放日志等相关的数据，当本申请实施例运用到具体产品或技术中时，需要获得用户许可或者同意，且相关数据的收集、使用和处理需要遵守相关国家和地区的相关法律法规和标准。
以上所揭露的仅为本申请较佳实施例而已，当然不能以此来限定本申请之权利范围，因此依本申请权利要求所作的等同变化，仍属本申请所涵盖的范围。

Claims (19)

  1. 一种业务数据处理方法，所述方法由电子设备执行，包括：
    获取N个已投放业务数据对应的投放日志；所述投放日志包括业务对象集合分别针对所述N个已投放业务数据的业务输出状态与业务转化状态；N为正整数；所述业务转化状态包括未转化状态；所述业务输出状态包括已输出状态；
    根据所述业务对象集合分别针对所述N个已投放业务数据的所述业务输出状态与所述业务转化状态，建立目标业务对象针对未转化业务数据的未转化浏览行为特征；所述目标业务对象针对所述未转化业务数据的所述业务输出状态为所述已输出状态、且所述业务转化状态为所述未转化状态；所述N个已投放业务数据包括所述未转化业务数据，所述业务对象集合包括所述目标业务对象；
    根据所述未转化浏览行为特征，确定所述目标业务对象与未转化物品共同对应的排序因子，根据所述排序因子，对包含有所述未转化物品的业务数据进行排序，得到序列业务数据，按照排序顺序从所述序列业务数据中选择目标业务数据，将所述目标业务数据投放至所述目标业务对象；所述未转化业务数据中包含所述未转化物品。
  2. 根据权利要求1所述的方法，其中，所述投放日志还包括所述业务对象集合分别针对所述N个已投放业务数据的记录时间戳；所述业务对象集合包括业务对象M i，所述N个已投放业务数据包括已投放业务数据G b，所述业务对象M i针对所述已投放业务数据G b的所述记录时间戳，是指所述业务对象M i针对所述已投放业务数据G b的状态变更操作的操作记录时间戳，所述状态变更操作用于指示所述业务对象M i针对所述已投放业务数据G b的所述业务输出状态由未输出状态转换为所述已输出状态；i、b均为正整数；
    所述根据所述业务对象集合分别针对所述N个已投放业务数据的所述业务输出状态与所述业务转化状态，建立目标业务对象针对未转化业务数据的未转化浏览行为特征，包括：
    获取所述业务对象M i的对象标识与所述已投放业务数据G b的业务标识；
    将所述业务对象M i的所述对象标识、所述已投放业务数据G b的所述业务标识、所述业务对象M i针对所述已投放业务数据G b的所述业务输出状态、所述业务对象M i针对所述已投放业务数据G b的所述业务转化状态、以及所述业务对象Mi针对所述已投放业务数据G b的所述记录时间戳所组成的数据组，确定为所述业务对象M i针对所述已投放业务数据G b的浏览行为特征；
    在所述业务对象集合针对所述N个已投放业务数据的所述浏览行为特征中，确定所述目标业务对象针对所述未转化业务数据的所述未转化浏览行为特征。
  3. 根据权利要求2所述的方法，其中，所述在所述业务对象集合针对所述N个已投放业务数据的所述浏览行为特征中，确定所述目标业务对象针对所述未转化业务数据的所述未转化浏览行为特征，包括：
    在所述业务对象集合针对所述N个已投放业务数据的所述浏览行为特征中，将所述业务输出状态为所述已输出状态、且所述业务转化状态为所述未转化状态的所述浏览行为特征，确定为所述未转化浏览行为特征；
    将所述未转化浏览行为特征中包含的所述对象标识所对应的业务对象，确定为所述目标业务对象，在包含所述目标业务对象的所述对象标识的所述未转化浏览行为特征中，将包含的所述业务标识所对应的已投放业务数据，确定为所述目标业务对象针对的所述未转化业务数据，以得到所述目标业务对象针对所述未转化业务数据的所述未转化浏览行为特征。
  4. 根据权利要求1-3任一项所述的方法，其中，所述根据所述未转化浏览行为特征，确定所述目标业务对象与未转化物品共同对应的排序因子，包括：
    对所述未转化业务数据进行业务识别，得到所述未转化业务数据与所述未转化物品之间的映射关系；所述映射关系用于指示所述未转化业务数据包含所述未转化物品；
    获取所述未转化物品的物品标识，将所述未转化浏览行为特征中包含的所述未转化业务数据的业务标识，替换为所述未转化物品的所述物品标识，将替换后的未转化浏览行为特征，确定为所述目标业务对象针对所述未转化物品的目标浏览行为特征；
    根据所述目标浏览行为特征，确定所述目标业务对象与所述未转化物品共同对应的所述排序因子。
  5. 根据权利要求4所述的方法，其中，包含有所述未转化物品的业务数据包括Q个目标未转化业务数据；所述Q个目标未转化业务数据中均包含同一未转化物品P x；Q为小于或等于N的正整数，x为正整数。
  6. 根据权利要求5所述的方法，其中，所述目标业务对象针对所述未转化物品的所述目标浏览行为特征，包括所述目标业务对象针对所述未转化物品P x的Q个目标浏览行为特征；
    所述根据所述目标浏览行为特征，确定所述目标业务对象与所述未转化物品共同对应的所述排序因子，包括：
    统计所述Q个目标浏览行为特征的特征总数量，将所述特征总数量确定为所述目标业务对象针对所述未转化物品P x的总输出次数；
    从所述Q个目标浏览行为特征分别包含的记录时间戳中，获取最早记录时间戳；
    根据当前时间戳、所述总输出次数、以及所述最早记录时间戳，确定所述目标业务对象与所述未转化物品P x共同对应的所述排序因子。
  7. 根据权利要求5所述的方法，其中，所述根据所述排序因子，对包含有所述未转化物品的业务数据进行排序，得到序列业务数据，包括：
    获取目标未转化业务数据R s针对所述业务对象集合的总投放量，所述目标未转化业务数据R s为所述Q个目标未转化业务数据中的任一个；
    根据所述总投放量、以及所述业务对象集合针对所述目标未转化业务数据R s的实时转化数量，确定所述目标未转化业务数据R s对应的实时投放转化率；
    根据所述排序因子、以及所述Q个目标未转化业务数据分别对应的所述实时投放转化率，对包含有所述未转化物品P x的业务数据进行排序，得到所述序列业务数据。
  8. 根据权利要求7所述的方法，其中，包含有所述未转化物品的业务数据还包括包含所述未转化物品P x的待投放业务数据；
    所述根据所述排序因子、以及所述Q个目标未转化业务数据分别对应的所述实时投放转化率，对包含有所述未转化物品P x的业务数据进行排序，得到所述序列业务数据，包括：
    获取所述待投放业务数据对应的预估投放转化率；
    按照所述Q个目标未转化业务数据分别对应的所述实时投放转化率，与所述待投放业务数据对应的所述预估投放转化率之间的大小顺序，对所述包含有所述未转化物品P x的业务数据进行排序，得到初始序列业务数据；
    从所述初始序列业务数据中，按序获取所述预估投放转化率最大的K个业务数据；K为小于W的正整数；W为所述Q个目标未转化业务数据与包含所述未转化物品P x的所述待投放业务数据的总数量；
    获取所述K个业务数据分别对应的期望转化单位投放资源、预估输出率、以及预估转化率；
    根据所述排序因子，以及所述K个业务数据分别对应的所述期望转化单位投放资源、所述预估输出率和所述预估转化率，对所述K个业务数据进行排序，得到所述序列业务数据。
  9. 根据权利要求8所述的方法，其中，业务数据R t为所述K个业务数据中的任一个，t为正整数；
    所述根据所述排序因子，以及所述K个业务数据分别对应的所述期望转化单位投放资源、所述预估输出率和所述预估转化率，对所述K个业务数据进行排序，得到所述序列业务数据，包括：
    根据所述业务数据R t对应的所述期望转化单位投放资源、所述预估输出率、以及所述预估转化率，确定所述业务数据R t对应的初始投放单位消耗资源；
    将所述排序因子与所述目标业务数据R t对应的所述初始投放单位消耗资源进行运算处理，得到所述目标业务数据R t对应的实时投放单位消耗资源；
    根据K个实时投放单位消耗资源之间的大小顺序，对所述K个业务数据进行排序，得到所述序列业务数据。
  10. 根据权利要求9所述的方法，其中，包含有所述未转化物品的业务数据还包括未输出业务数据；所述目标业务对象针对所述未输出业务数据的所述业务输出状态为未输出状态；
    所述根据K个实时投放单位消耗资源之间的大小顺序，对所述K个业务数据进行排序，得到所述序列业务数据，包括：
    获取所述未输出业务数据对应的所述期望转化单位投放资源、所述预估输出率、以及所述预估转化率；
    根据所述未输出业务数据对应的所述期望转化单位投放资源、所述预估输出率、以及所述预估转化率，确定所述未输出业务数据对应的所述初始投放单位消耗资源；
    按照所述未输出业务数据对应的所述初始投放单位消耗资源，与所述K个实时投放单位消耗资源之间的大小顺序，对所述K个业务数据与所述未输出业务数据进行排序，得到所述序列业务数据。
  11. 根据权利要求1所述的方法，其中，所述方法还包括：
    在目标时间段内，获取所述目标业务对象针对所述目标业务数据的浏览反馈数据；
    根据所述浏览反馈数据，对所述目标业务对象进行相似业务数据投放处理。
  12. 根据权利要求11所述的方法，其中，所述浏览反馈数据包括以下至少一项：所述目标业务对象针对所述目标业务数据的转化反馈行为数据，所述目标业务对象针对所述目标业务数据的负向反馈行为数据，所述目标业务对象针对所述目标业务数据的目标浏览时长。
  13. 根据权利要求11或12所述的方法，其中，所述浏览反馈数据包括转化反馈行为数据；
    所述根据所述浏览反馈数据，对所述目标业务对象进行相似业务数据投放处理，包括：
    从所述目标业务数据中，获取所述转化反馈行为数据对应的转化业务数据；
    获取所述转化业务数据中所包括的转化物品，获取所述转化物品的第一相似物品；
    将所述第一相似物品所属的业务数据作为目标相似业务数据，将所述目标相似业务数据投放至所述目标业务对象。
  14. 根据权利要求11或12所述的方法，其中，所述浏览反馈数据包括负向反馈行为数据；
    所述根据所述浏览反馈数据，对所述目标业务对象进行相似业务数据投放处理，包括：
    从所述目标业务数据中，获取所述负向反馈行为数据对应的负向反馈业务数据；
    获取所述负向反馈业务数据中所包括的负向反馈物品，获取所述负向反馈物品的第二相似物品；
    将所述第二相似物品所属的业务数据作为目标相似业务数据，在针对所述目标业务对象的业务数据投放过程中，对所述目标相似业务数据进行过滤。
  15. 根据权利要求11或12所述的方法，其中，所述浏览反馈数据包括目标浏览时长；
    所述根据所述浏览反馈数据，对所述目标业务对象进行相似业务数据投放处理，包括：
    获取所述目标业务对象针对所述目标业务数据的历史平均浏览时长；
    当所述目标浏览时长大于所述历史平均浏览时长时，将所述目标业务数据中包括的业务物品确定为潜在转化物品，将所述潜在转化物品的第三相似物品所属的业务数据作为目标相似业务数据，将所述目标相似业务数据投放至所述目标业务对象。
  16. 一种业务数据处理装置，包括：
    日志获取模块，配置为获取N个已投放业务数据对应的投放日志；所述投放日志包括业务对象集合分别针对所述N个已投放业务数据的业务输出状态与业务转化状态；N为正整数；所述业务转化状态包括未转化状态；所述业务输出状态包括已输出状态；
    特征建立模块，配置为根据所述业务对象集合分别针对所述N个已投放业务数据的所述业务输出状态与所述业务转化状态，建立目标业务对象针对未转化业务数据的未转化浏览行为特征；所述目标业务对象针对所述未转化业务数据的所述业务输出状态为所述已输出状态、且所述业务转化状态为所述未转化状态；所述N个已投放业务数据包括所述未转化业务数据，所述业务对象集合包括所述目标业务对象；
    排序因子确定模块，配置为根据所述未转化浏览行为特征，确定所述目标业务对象与未转化物品共同对应的排序因子；所述未转化业务数据中包含所述未转化物品；
    数据排序模块，配置为根据所述排序因子，对包含有所述未转化物品的业务数据进行排序，得到序列业务数据；
    数据投放模块，配置为按照排序顺序从所述序列业务数据中选择目标业务数据，将所述目标业务数据投放至所述目标业务对象。
  17. 一种电子设备，包括：处理器、存储器以及网络接口；
    所述处理器与所述存储器、所述网络接口相连，其中，所述网络接口用于提供网络通信功能，所述存储器用于存储程序代码，所述处理器用于调用所述程序代码，以使所述电子设备执行权利要求1-15任一项所述的方法。
  18. 一种计算机可读存储介质，所述计算机可读存储介质中存储有计算机程序，所述计算机程序适于由处理器加载并执行权利要求1-15任一项所述的方法。
  19. 一种计算机程序产品，包括计算机程序或指令，所述计算机程序或指令被处理器执行时，实现权利要求1-15任一项所述的方法。
PCT/CN2022/099533 2021-08-11 2022-06-17 一种业务数据处理方法、装置、电子设备、计算机可读存储介质及计算机程序产品 WO2023016085A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US18/316,081 US20230281662A1 (en) 2021-08-11 2023-05-11 Service data processing method and apparatus, electronic device, computer-readable storage medium and computer program product

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202110919651.0 2021-08-11
CN202110919651.0A CN115705588A (zh) 2021-08-11 2021-08-11 一种业务数据处理方法、设备及可读存储介质

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US18/316,081 Continuation US20230281662A1 (en) 2021-08-11 2023-05-11 Service data processing method and apparatus, electronic device, computer-readable storage medium and computer program product

Publications (1)

Publication Number Publication Date
WO2023016085A1 true WO2023016085A1 (zh) 2023-02-16

Family

ID=85180116

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/099533 WO2023016085A1 (zh) 2021-08-11 2022-06-17 一种业务数据处理方法、装置、电子设备、计算机可读存储介质及计算机程序产品

Country Status (3)

Country Link
US (1) US20230281662A1 (zh)
CN (1) CN115705588A (zh)
WO (1) WO2023016085A1 (zh)

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2017134468A (ja) * 2016-01-25 2017-08-03 ヤフー株式会社 広告装置、調整方法および調整プログラム
CN109087178A (zh) * 2018-08-28 2018-12-25 清华大学 商品推荐方法和装置
US20190034961A1 (en) * 2017-07-26 2019-01-31 Accelerize Inc Method for targeting electronic advertising by data encoding and prediction for sequential data machine learning models
CN110516163A (zh) * 2019-07-09 2019-11-29 广州品唯软件有限公司 一种基于用户行为数据的商品排序方法及***
CN111667312A (zh) * 2020-06-08 2020-09-15 腾讯科技（深圳）有限公司 一种广告投放的方法、相关装置、设备以及存储介质
CN111667311A (zh) * 2020-06-08 2020-09-15 腾讯科技（深圳）有限公司 一种广告投放的方法、相关装置、设备以及存储介质

Family Cites Families (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9858341B2 (en) * 2006-08-02 2018-01-02 Jason Frankovitz Method and apparatus for remotely monitoring a social website
WO2009137507A2 (en) * 2008-05-05 2009-11-12 Berman, Joel, F. Preservation of scores of the quality of traffic to network sites across clients and over time
US8655695B1 (en) * 2010-05-07 2014-02-18 Aol Advertising Inc. Systems and methods for generating expanded user segments
WO2014107451A1 (en) * 2013-01-02 2014-07-10 Shared2You, Inc. User-based analysis of advertisement pools
FR3010206A1 (fr) * 2013-09-04 2015-03-06 Kila Systems Procede de gestion d'une base de donnees utilisateurs sur une plate-forme de contenus multimedias
US11379883B2 (en) * 2019-08-09 2022-07-05 SOCI, Inc. Systems, devices, and methods for dynamically generating, distributing, and managing online communications

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2017134468A (ja) * 2016-01-25 2017-08-03 ヤフー株式会社 広告装置、調整方法および調整プログラム
US20190034961A1 (en) * 2017-07-26 2019-01-31 Accelerize Inc Method for targeting electronic advertising by data encoding and prediction for sequential data machine learning models
CN109087178A (zh) * 2018-08-28 2018-12-25 清华大学 商品推荐方法和装置
CN110516163A (zh) * 2019-07-09 2019-11-29 广州品唯软件有限公司 一种基于用户行为数据的商品排序方法及***
CN111667312A (zh) * 2020-06-08 2020-09-15 腾讯科技（深圳）有限公司 一种广告投放的方法、相关装置、设备以及存储介质
CN111667311A (zh) * 2020-06-08 2020-09-15 腾讯科技（深圳）有限公司 一种广告投放的方法、相关装置、设备以及存储介质

Also Published As

Publication number Publication date
CN115705588A (zh) 2023-02-17
US20230281662A1 (en) 2023-09-07

Similar Documents

Publication Publication Date Title
US20160162924A1 (en) Track user activities and in exchange provide points to use for various activities including advertising contents to targeted users of network
CN107370780B (zh) 基于互联网的媒体推送方法、装置和***
WO2017128933A1 (zh) 一种针对作品的处理方法、装置及***
US11030641B2 (en) Method and system for online advertising
WO2009085524A2 (en) Using product and social network data to improve online advertising
CN109831501B (zh) 基于区块链的信息分发方法和***
US20170046749A1 (en) One-click promotional advertising
US20190108545A1 (en) Automatically detecting and modifying an exploit digital advertising campaign
US20190043094A1 (en) Presenting a content item to an online system user including content dynamically generated by the online system for the user
WO2022237477A1 (zh) 一种数据校准方法、装置、计算机设备以及可读存储介质
US20170213238A1 (en) Evaluating performance of content item presentation by modifying parameters determining whether to present content item to users of an online system
CN111865753B (zh) 媒体信息的参数确定方法和装置、存储介质、电子装置
CN116383480A (zh) 通话场景下媒体推荐的方法、装置、电子设备及存储介质
US20120041850A1 (en) Incentivizing content-receivers in social networks
CN111582899A (zh) 一种展示链接效果的量化方法、装置、***及存储介质
JP2022096666A (ja) 収益分配装置、収益分配方法及び収益分配システム
US20160092852A1 (en) Allocation and distribution of payment for podcast services
WO2023016085A1 (zh) 一种业务数据处理方法、装置、电子设备、计算机可读存储介质及计算机程序产品
CN112734502B (zh) 一种多媒体信息定向投放的测试方法、装置和电子设备
US10963921B2 (en) Presenting content to an online system user assigned to a stage of a classification scheme and determining a value associated with an advancement of the user to a succeeding stage
CN109379722A (zh) 一种聊天广告的方法与***
CN110245970A (zh) 一种聊天广告的方法与***
CN115082095A (zh) 转化率预测模型训练方法、装置、电子设备及存储介质
KR20080095122A (ko) 신규 회원등록에 따른 인센티브 증가에 의한 회원관리시스템 및 그 방법
US20130332288A1 (en) Advertising media and method of advertising

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 22855070

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE