WO2024027125A1 - 对象推荐方法、装置、电子设备和存储介质 - Google Patents

对象推荐方法、装置、电子设备和存储介质 Download PDF

Info

Publication number
WO2024027125A1
WO2024027125A1 PCT/CN2023/075417 CN2023075417W WO2024027125A1 WO 2024027125 A1 WO2024027125 A1 WO 2024027125A1 CN 2023075417 W CN2023075417 W CN 2023075417W WO 2024027125 A1 WO2024027125 A1 WO 2024027125A1
Authority
WO
WIPO (PCT)
Prior art keywords
description
target
search
text
representation vector
Prior art date
Application number
PCT/CN2023/075417
Other languages
English (en)
French (fr)
Inventor
王欢
王培建
Original Assignee
百度在线网络技术（北京）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 百度在线网络技术（北京）有限公司 filed Critical 百度在线网络技术（北京）有限公司
Publication of WO2024027125A1 publication Critical patent/WO2024027125A1/zh

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90—Details of database functions independent of the retrieved data types
    • G06F16/95—Retrieval from the web
    • G06F16/953—Querying, e.g. by the use of web search engines
    • G06F16/9535—Search customisation based on user profiles and personalisation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
    • G06F16/33—Querying
    • G06F16/3331—Query processing
    • G06F16/334—Query execution
    • G06F16/3344—Query execution using natural language analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F40/00—Handling natural language data
    • G06F40/20—Natural language analysis
    • G06F40/205—Parsing
    • G06F40/216—Parsing using statistical methods
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F40/00—Handling natural language data
    • G06F40/20—Natural language analysis
    • G06F40/279—Recognition of textual entities
    • G06F40/289—Phrasal analysis, e.g. finite state techniques or chunking
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F40/00—Handling natural language data
    • G06F40/30—Semantic analysis
    • 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
    • G06Q30/06—Buying, selling or leasing transactions
    • G06Q30/0601—Electronic shopping [e-shopping]
    • G06Q30/0631—Item recommendations

Definitions

  • the present disclosure relates to the field of artificial intelligence technology, in particular to the field of recommendation technology based on artificial intelligence, and specifically to an object recommendation method, device, electronic device, computer-readable storage medium and computer program product.
  • Artificial intelligence is the study of using computers to simulate certain human thinking processes and intelligent behaviors (such as learning, reasoning, thinking, planning, etc.). It has both hardware-level technology and software-level technology. Artificial intelligence hardware technology generally includes technologies such as sensors, dedicated artificial intelligence chips, cloud computing, distributed storage, big data processing, etc.; artificial intelligence software technology mainly includes computer vision technology, speech recognition technology, natural language processing technology, and machine learning/depth Learning, big data processing technology, knowledge graph technology and other major directions.
  • object recommendation based on artificial intelligence predicts the user's preferences for objects and recommends objects that match the user's preferences.
  • the present disclosure provides an object recommendation method, device, electronic device, computer-readable storage medium and computer program product.
  • an object recommendation method including: obtaining an object to be recommended, the object having a corresponding description text; obtaining a plurality of target descriptions among a plurality of description participles included in the description text Word segmentation, the plurality of target description word segments are used to distinguish the description text from other description texts corresponding to other objects; based on the plurality of target description word segments, an object representation vector of the object is obtained; and based on the object Representation vector that recommends the object.
  • an object recommendation device including: an object acquisition unit configured to obtain an object to be recommended, the object having a corresponding description text; a target description segmentation acquisition unit configured with Obtaining a plurality of target description participles among a plurality of description participles included in the description text, the plurality of target description participles being used to distinguish the description text from other description texts corresponding to other objects; object representation vector acquisition a unit configured to obtain an object representation vector of the object based on the plurality of target description word segments; and a recommendation unit configured to recommend the object based on the object representation vector.
  • an electronic device including: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores information that can be executed by the at least one processor.
  • the instructions are executed by the at least one processor to enable the at least one processor to perform the method according to the embodiment of the present disclosure.
  • a non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to cause the computer to perform a method according to an embodiment of the present disclosure.
  • a computer program product comprising a computer program, wherein the computer program, when executed by a processor, implements a method according to an embodiment of the present disclosure.
  • the accuracy of recommended objects for users can be improved.
  • FIG. 1 illustrates a schematic diagram of an exemplary system in which various methods described herein may be implemented in accordance with embodiments of the present disclosure
  • Figure 2 shows a flow chart of an object recommendation method according to an embodiment of the present disclosure
  • FIG. 3 shows a flowchart of a process of obtaining a plurality of target description participles among a plurality of description participles included in a description text in an object recommendation method according to an embodiment of the present disclosure
  • Figure 4 shows a flow chart of a process of obtaining an object representation vector of an object based on multiple target description word segmentations in an object recommendation method according to an embodiment of the present disclosure
  • FIG. 5 shows a flow chart of a process of obtaining an object representation vector based on a semantic representation vector of each target description segment among multiple target description segmentations in an object recommendation method according to an embodiment of the present disclosure
  • Figure 6 shows a flow chart of an object recommendation method according to an embodiment of the present disclosure
  • FIG. 7 shows a flowchart of a process of obtaining multiple target search word segments among multiple search word segments included in the search text in an object recommendation method according to an embodiment of the present disclosure
  • FIG. 8 shows a flow chart of a process of obtaining a search text representation vector of a search text based on multiple target search word segmentations in an object recommendation method according to an embodiment of the present disclosure
  • Figure 9 shows a structural block diagram of an object recommendation device according to an embodiment of the present disclosure.
  • FIG. 10 shows a structural block diagram of an exemplary electronic device that can be used to implement embodiments of the present disclosure.
  • first”, “second”, etc. to describe various elements is not intended to limit the positional relationship, timing relationship, or importance relationship of these elements. Such terms are only used for Distinguish one feature from another.
  • first element and the second element may refer to the same instance of the element, and in some cases, based on contextual description, they may refer to different instances.
  • FIG. 1 shows a schematic diagram of an exemplary system 100 in which various methods and apparatus described herein may be implemented in accordance with embodiments of the present disclosure.
  • the system 100 includes one or more client devices 101 , 102 , 103 , 104 , 105 , and 106 , a server 120 , and one or more communication networks coupling the one or more client devices to the server 120 110.
  • Client devices 101, 102, 103, 104, 105, and 106 may be configured to execute one or more applications.
  • the server 120 may run one or more services or software applications that enable execution of the object recommendation method according to the present disclosure.
  • server 120 may also provide other services or software applications, which may include non-virtual environments and virtual environments.
  • these services may be provided as web-based services or cloud services, such as under a Software as a Service (SaaS) model to users of client devices 101, 102, 103, 104, 105, and/or 106 .
  • SaaS Software as a Service
  • server 120 may include one or more components that implement the functions performed by server 120 . These components may include software components, hardware components, or combinations thereof that are executable by one or more processors. Users operating client devices 101, 102, 103, 104, 105, and/or 106 may, in turn, utilize one or more client applications to interact with server 120 to utilize services provided by these components. It should be understood that a variety of different system configurations are possible, which may differ from system 100 . Accordingly, Figure 1 is one example of a system for implementing the various methods described herein and is not intended to be limiting.
  • a user may use client devices 101, 102, 103, 104, 105, and/or 106 to receive one or more objects recommended to the user.
  • the client device may provide an interface that enables a user of the client device to interact with the client device.
  • the client device can also output information to the user via the interface.
  • FIG. 1 depicts only six client devices, those skilled in the art will understand that the present disclosure can support any number of client devices.
  • Client devices 101, 102, 103, 104, 105, and/or 106 may include various types of computer devices, such as portable handheld devices, general purpose computers (such as personal computers and laptop computers), workstation computers, wearable devices, Smart screen equipment, self-service terminal equipment, service robots, game systems, thin clients, various messaging equipment, sensors or other sensing equipment, etc.
  • These computer devices can run various types and versions of software applications and operating systems, such as MICROSOFT Windows, APPLE iOS, UNIX-like operating systems, Linux or Linux-like operating systems (such as GOOGLE Chrome OS); or include various mobile operating systems , such as MICROSOFT Windows Mobile OS, iOS, Windows Phone, Android.
  • Portable handheld devices may include cellular phones, smart phones, tablet computers, personal digital assistants (PDAs), and the like.
  • Wearable devices may include head-mounted displays (such as smart glasses) and other devices.
  • Gaming systems may include various handheld gaming devices, Internet-enabled gaming devices, and the like.
  • the client device is capable of executing a variety of different applications, such as various Internet-related applications, communication applications (such as email applications), Short Message Service (SMS) applications, and can use various communication protocols.
  • Network 110 may be any type of network known to those skilled in the art that may support data communications using any of a variety of available protocols (including, but not limited to, TCP/IP, SNA, IPX, etc.).
  • one or more networks 110 may be a local area network (LAN), an Ethernet-based network, a token ring, a wide area network (WAN), the Internet, a virtual network, a virtual private network (VPN), an intranet, an extranet, Blockchain networks, Public Switched Telephone Network (PSTN), infrared networks, wireless networks (e.g. Bluetooth, WIFI) and/or any combination of these and/or other networks.
  • LAN local area network
  • Ethernet-based network a token ring
  • WAN wide area network
  • VPN virtual private network
  • PSTN Public Switched Telephone Network
  • WIFI wireless networks
  • Server 120 may include one or more general purpose computers, special purpose server computers (eg, PC (Personal Computer) servers, UNIX servers, midrange servers), blade servers, mainframe computers, server clusters, or any other suitable arrangement and/or combination .
  • Server 120 may include one or more virtual machines running a virtual operating system, or other computing architecture involving virtualization (e.g., one of the logical storage devices that may be virtualized to maintain the server's virtual storage device). or multiple flexible pools).
  • server 120 may run one or more services or software applications that provide the functionality described below.
  • Computing units in server 120 may run one or more operating systems, including any of the operating systems described above, as well as any commercially available server operating system.
  • Server 120 may also run any of a variety of additional server applications and/or middle-tier applications, including HTTP servers, FTP servers, CGI servers, JAVA servers, database servers, and the like.
  • server 120 may include one or more applications to analyze and incorporate data feeds and/or event updates received from users of client devices 101, 102, 103, 104, 105, and/or 106.
  • Server 120 may also include one or more applications to display data feeds and/or real-time events via one or more display devices of client devices 101 , 102 , 103 , 104 , 105 , and/or 106 .
  • the server 120 may be a server of a distributed system, or a server combined with a blockchain.
  • the server 120 may also be a cloud server, or an intelligent cloud computing server or intelligent cloud host with artificial intelligence technology.
  • Cloud server is a host product in the cloud computing service system to solve the shortcomings of difficult management and weak business scalability in traditional physical host and virtual private server (VPS) services.
  • System 100 may also include one or more databases 130.
  • these databases may be used to store data and other information.
  • databases 130 may be used to store information such as audio files and video files.
  • Database 130 may reside in various locations.
  • a database used by server 120 may be local to server 120, or may be remote from server 120 and may communicate with server 120 via a network-based or dedicated connection.
  • Database 130 may be of different types.
  • the database used by server 120 may be, for example, a relational database.
  • One or more of these databases may store, update, and retrieve data to and from the database in response to commands.
  • one or more of databases 130 may also be used by applications to store application data.
  • the database used by the application can be different types of databases such as key-value repositories, object repositories or regular repositories backed by a file system.
  • the system 100 of Figure 1 may be configured and operated in various ways to enable the application of the various methods and apparatus described in accordance with the present disclosure.
  • an object recommendation method 200 includes:
  • Step S210 Obtain the object to be recommended, and the object has corresponding description text
  • Step S220 Obtain a plurality of target description participles among a plurality of description participles included in the description text, and the plurality of target description participles are used to distinguish the description text from other description texts corresponding to other objects;
  • Step S230 Obtain the object representation vector of the object based on the plurality of target description word segments.
  • Step S240 Recommend the object based on the object representation vector.
  • a representation vector of an object is often obtained based on the description text, and based on the representation vector of the object, it is determined whether to recommend the object to the user. Since the description text often contains words that are irrelevant to the object and cannot distinguish the object from other objects, for example, words describing the application scenarios of the object, the obtained representation vector of the object is not representative; based on the representation vector
  • the objects recommended for users are often far from their expectations, especially in the process of matching the representation vector with the user's search request, because the user's search request is often relatively concise. , often cannot match the representation vector that exactly matches it, making the objects recommended for users inaccurate.
  • the plurality of target description participles can distinguish the description text of the object from other description texts of other objects.
  • the object representation vector of the object is obtained, so that the obtained object representation vector can distinguish the object from other objects, achieve focused representation of the object, and then make recommendations based on the object representation vector.
  • accurate matching can be performed to improve the accuracy of objects recommended to users.
  • the objects to be recommended can be any information, resources, etc. that exist in the form of electronic data, such as videos, articles, products, etc.
  • the description text of the object when the object is a video, the description text of the object may be the title, subtitles, tags, etc. of the video; when the object is a commodity, the description text may be the title of the commodity; when the object is a violation, the description text may be It is the title of the article or the content text of the article.
  • the object includes a commodity
  • the description text includes a title of the commodity
  • obtaining a plurality of target description participles among a plurality of description participles included in the description text includes:
  • Step S310 Segment the description text to obtain the multiple description participles
  • Step S320 Obtain a first score corresponding to each description participle of the plurality of description participles, the first score indicating the possibility of distinguishing the description text from other description texts corresponding to other objects based on the corresponding description participle. ;as well as
  • Step S330 Obtain the plurality of target description participles based on the plurality of first scores corresponding to the plurality of description participles.
  • the first score indicating the possibility of distinguishing the description text from other description texts corresponding to other objects based on the corresponding description participle, and based on the first score for each description participle Obtain multiple target description participles and make the obtained multiple target description participles accurate.
  • each description participle is input to a trained scoring model to obtain a first score corresponding to each description participle.
  • the first score corresponding to each description participle in the plurality of description participles is obtained through term frequency-inverse document frequency (TF-IDF).
  • TF-IDF term frequency-inverse document frequency
  • the description text set is obtained by obtaining the description text of each object in the object set.
  • the first score S of the description participle x in the i-th description text I is obtained through formula (1):
  • x i is the number of times the description participle x appears in the description text I
  • I n is the total number of times each description participle appears in the description text I
  • N is the total number of texts in the description text collection
  • N x is the number of times the description participle x appears in the description text I. The total number of description texts.
  • the weight of each description participle in the description text for dividing the description text and other description texts can be obtained (i.e., the first score S).
  • a description word segment with a first score greater than a preset score threshold among multiple description word segments is determined as a target description word segment to obtain the plurality of target description word segments.
  • the plurality of target description participles include a preset number of description participles among the plurality of description participles, and the first score of each description participle in the preset number of description participles is higher than Other description participles in the plurality of description participles, the other description participles are different from each description participle in the preset number of description participles.
  • the multiple target description participles are the preset number of description participles with the first larger score among the multiple description participles, so that the obtained target description participle is the more representative description participle among the multiple description participles, and then the multiple target description participles are based on the multiple description participles.
  • the object representation vector of the object obtained by the target description segmentation is more accurate.
  • the object representation vector of the object is obtained directly based on the word vector of each of the plurality of target description participles.
  • obtaining the object representation vector of the object includes:
  • Step S410 Obtain a semantic representation vector of each target description segment among the plurality of target description segments, the semantic representation vector being related to the position of the target description segment in the description text;
  • Step S420 Obtain the object representation vector based on the semantic representation vector of each target description word segment among the plurality of target description word segments.
  • the semantic representation vector is related to the position of the target description segment in the description text, so that the semantic representation vector of each target description segment is related to the semantics of the description text, and it is possible to mine Describe the similar information between the deep semantics in the text, and then make the object representation vector obtained based on each semantic representation vector based on the target description word segmentation contain the similar information between the deep semantics in the description text, so that the object representation vector can represent the object precise.
  • a semantic vector representation of each of the multiple target description word segments is obtained.
  • a word sequence composed of the multiple description participles according to their order in the description text is input into the BERT model to obtain the semantic representation vector of each description participle. , which includes the semantic representation vector of each target description participle in the plurality of target description participles.
  • the object representation vector of the object is obtained by directly adding the plurality of semantic representation vectors of the plurality of target description word segments.
  • the object representation vector is obtained based on the semantic representation vector and the first score of each of the plurality of target description segmentations.
  • obtaining the object representation vector includes:
  • Step S510 Normalize the first score of each target description word segment among the plurality of target description word segments to obtain a weighted score of the description word segment;
  • Step S520 perform weighting processing on the semantic object representation vector of the target description word segment based on the weighted score of each target description word segment among the plurality of target description word segments, to obtain the weighted vector of the target description word segment;
  • Step S530 Obtain the object representation vector based on the weighted vector of each target description word segment among the plurality of target description word segments.
  • the object representation vector By weighting the corresponding semantic representation vector based on the weighted score obtained after normalizing the first score of each target description segment, and obtaining the object representation vector based on the weighted vector obtained after the weighting process, so that the obtained
  • the object representation vector also includes the distinction between the importance of each target description segment to the description text (the possibility of distinguishing the description text from other description texts based on the target description segment), further improving the obtained object representation vector. accuracy.
  • the object representation vector is obtained by directly adding multiple weighted vectors corresponding to multiple target description segments.
  • the user's user representation vector is also obtained, the similarity between the object representation vector and the user representation vector is calculated, and based on the similarity, it is determined whether to recommend the object to the user.
  • the object recommendation method according to some embodiments of the present disclosure further includes:
  • Step S610 Obtain the user's search text
  • Step S620 Obtain a plurality of target search word segments among a plurality of search word segments included in the search text, and the plurality of target search word segments are used to distinguish the search text from other search texts;
  • Step S630 Based on the plurality of target search word segments, obtain a search text representation vector of the search text; and wherein recommending the object based on the object representation vector includes:
  • the process of recommending objects to the user based on the user's search text matching between text and text is achieved. Due to the description text of the object and the user's search text, multiple items in the text are obtained respectively. Multiple target segmentations in the word segmentation that can distinguish it from other corresponding texts, and corresponding representation vectors are obtained based on the corresponding multiple target segmentations. Among them, the object representation vector is obtained based on multiple target description segmentations of the description text, and the object representation vector is obtained based on the search text.
  • the multiple target search word segmentation obtains the search text representation vector, which improves the accuracy of characterizing the object and the search text, so that when matching based on the search text and the description text of the object, the matching results are more accurate, and thus the objects recommended for the user can be improved accuracy.
  • obtaining a plurality of target search word segments among a plurality of search word segments included in the search text includes:
  • Step S710 Segment the search text to obtain the multiple search word segments
  • Step S720 Obtain a second score corresponding to each search word segmentation in the plurality of search word segments, the second score indicating the possibility of distinguishing the search text from other search texts based on the corresponding search word segmentation;
  • Step S730 Obtain the plurality of target search word segments based on the plurality of second scores corresponding to the plurality of search word segments.
  • Multiple target searches are obtained by obtaining a second score for each search participle in the search text that indicates the likelihood of distinguishing the search text from other search texts based on the corresponding search participle.
  • Word segmentation enables accurate word segmentation of multiple target searches.
  • each search word segment is input to a trained scoring model to obtain a second score corresponding to each search word segment.
  • the second score corresponding to each search word segmentation in the plurality of search word segments is obtained through term frequency-inverse document frequency (TF-IDF).
  • TF-IDF term frequency-inverse document frequency
  • obtaining the search text representation vector of the search text includes:
  • Step S810 Obtain a semantic representation vector of each target search word segment among the plurality of target search word segments, where the semantic representation vector is related to the position of the target search word segment in the search text;
  • Step S820 Based on the second score corresponding to each target search word segment among the plurality of target search word segments, obtain a weighted score corresponding to the target search word segment;
  • Step S830 Obtain the search text representation vector based on the semantic representation vector and weighted score of each target search segment among the plurality of target search segments.
  • the semantic representation vector is related to the position of the target search word segment in the search text, so that the semantic representation vector of each target search word segment is related to the semantics of the search text, and the search text can be mined
  • the similar information between the deep semantics in the search text, and then the search text representation vector obtained based on each semantic representation vector based on the target search word segmentation contains the similar information between the deep semantics in the search text, so that the search text representation vector has an important influence on the search text.
  • the characterization is accurate.
  • the corresponding semantic representation vector is weighted based on the weighted score obtained after normalizing the second score of each target search word segmentation, and the search text representation vector is obtained based on the weighted vector obtained after the weighting process,
  • the obtained search text representation vector also includes the distinction between the importance of each target search segment to the search text (the possibility of distinguishing the search text from other search texts based on the target search segment), further improving the obtained Accuracy in searching text representation vectors.
  • a weighted vector is obtained by multiplying the semantic representation vector of each target search word segment and the weighted score; and a search text representation vector is obtained by adding multiple weighted vectors corresponding to multiple target search word segments.
  • the search text representation vector after obtaining the search text representation vector, calculate the similarity between the search text representation vector and the object representation vector, and determine whether to recommend the object to the user based on the similarity.
  • an object recommendation device is also provided.
  • the device 900 includes: an object acquisition unit 910 configured to obtain an object to be recommended, the object having There is corresponding description text; the target description participle acquisition unit 920 is configured to obtain a plurality of target description participles among a plurality of description participles included in the description text, and the plurality of target description participles are used to obtain the description
  • the text is distinguished from other description texts corresponding to other objects;
  • the object representation vector acquisition unit 930 is configured to obtain the object representation vector of the object based on the plurality of target description word segmentations; and the recommendation unit 940 is configured to Based on the object representation vector, the object is recommended.
  • the target description word segmentation acquisition unit 920 includes: a word segmentation unit configured to segment the description text to obtain the plurality of description word segments; a first score calculation unit configured For obtaining a first score corresponding to each description participle in the plurality of description participles, the first score indicating the possibility of distinguishing the description text from other description texts corresponding to other objects based on the corresponding description participle; and a target description participle acquisition subunit configured to obtain the plurality of target description participles based on a plurality of first scores corresponding to the plurality of description participles.
  • the plurality of target description participles include a preset number of description participles among the plurality of description participles, and the first score of each description participle in the preset number of description participles is higher than Other description participles in the plurality of description participles, the other description participles are different from each description participle in the preset number of description participles.
  • the object representation vector acquisition unit 930 includes: a semantic representation vector acquisition unit configured to acquire a semantic representation vector for each target description segment among the plurality of target description segmentations, the semantic representation vector related to the position of the target description participle in the description text; and an object representation vector acquisition subunit configured to obtain the semantic representation vector based on each target description participle in the plurality of target description participles.
  • a semantic representation vector acquisition unit configured to acquire a semantic representation vector for each target description segment among the plurality of target description segmentations, the semantic representation vector related to the position of the target description participle in the description text
  • an object representation vector acquisition subunit configured to obtain the semantic representation vector based on each target description participle in the plurality of target description participles.
  • the object representation vector acquisition subunit includes: a normalization unit configured to normalize the first score of each target description participle in the plurality of target description participles, To obtain a weighted score of the description participle; the weighting unit is configured to weight the semantic object representation vector of the target description participle based on the weighted score of each target description participle in the plurality of target description participles, to obtain The weighted vector of the target description participles; and the first acquisition subunit is configured to obtain the object representation vector based on the weighted vector of each target description participle in the plurality of target description participles.
  • the method further includes: a search text acquisition unit configured to obtain the user's search text; a target search word segmentation acquisition unit configured to obtain multiple search word segments included in the search text. Target search word segmentation, the plurality of target search word segments are used to distinguish the search text from other search texts; and a search text representation vector acquisition unit configured to obtain the search based on the plurality of target search word segments a search text representation vector of the text; and wherein the recommendation unit 930 includes: a determination unit configured to determine whether to recommend the object to the user based on the search text representation vector and the object representation vector.
  • obtaining a plurality of target search word segments among a plurality of search word segments included in the search text includes: segmenting the search text to obtain the multiple search word segments; obtaining the a second score corresponding to each of the plurality of search word segments, the second score indicating a possibility of distinguishing the search text from other search texts based on the corresponding search word segment; and a corresponding score based on the plurality of search word segments A plurality of second scores are obtained to obtain the plurality of target search word segments.
  • obtaining a search text representation vector of the search text based on the plurality of target search word segments includes: obtaining a semantic representation vector of each target search word segmentation in the plurality of target search word segments, the The semantic representation vector is related to the position of the target search word segment in the search text; based on the second score corresponding to each target search word segment in the plurality of target search word segments, a weighted score corresponding to the target search word segment is obtained; and The search text representation vector is obtained based on a semantic representation vector and a weighted score of each target search segment among the plurality of target search segments.
  • the collection, storage, use, processing, transmission, provision and disclosure of user personal information are in compliance with relevant laws and regulations and do not violate public order and good customs.
  • an electronic device a readable storage medium, and a computer program product are also provided.
  • Electronic devices are intended to refer to various forms of digital electronic computing equipment, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. Electronic devices may also represent various forms of mobile devices, such as personal Word processors, cellular phones, smart phones, wearable devices, and other similar computing devices.
  • the components shown herein, their connections and relationships, and their functions are examples only and are not intended to limit implementations of the disclosure described and/or claimed herein.
  • the electronic device 1000 includes a computing unit 1001 that can perform calculations according to a computer program stored in a read-only memory (ROM) 1002 or loaded from a storage unit 1008 into a random access memory (RAM) 1003 . Perform various appropriate actions and processing.
  • RAM random access memory
  • various programs and data required for the operation of the electronic device 1000 can also be stored.
  • Computing unit 1001, ROM 1002 and RAM 1003 are connected to each other via bus 1004.
  • An input/output (I/O) interface 1005 is also connected to bus 1004.
  • the input unit 1006 may be any type of device capable of inputting information to the electronic device 1000, the input unit 1006 may receive input numeric or character information, and generate key signal input related to user settings and/or function control of the electronic device, and This may include, but is not limited to, a mouse, keyboard, touch screen, trackpad, trackball, joystick, microphone, and/or remote control.
  • Output unit 1007 may be any type of device capable of presenting information, and may include, but is not limited to, a display, speakers, video/audio output terminal, vibrator, and/or printer.
  • the storage unit 1008 may include, but is not limited to, magnetic disks and optical disks.
  • the communication unit 1009 allows the electronic device 1000 to exchange information/data with other devices through a computer network such as the Internet and/or various telecommunications networks, and may include, but is not limited to, a modem, a network card, an infrared communication device, a wireless communication transceiver and/or a chip Groups such as BluetoothTM devices, 802.11 devices, WiFi devices, WiMax devices, cellular communications devices, and/or the like.
  • Computing unit 1001 may be a variety of general and/or special purpose processing components having processing and computing capabilities. Some examples of the computing unit 1001 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, digital signal processing processor (DSP), and any appropriate processor, controller, microcontroller, etc.
  • the computing unit 1001 performs various methods and processes described above, such as method 200.
  • method 200 may be implemented as a computer software program tangibly embodied in a machine-readable medium, such as storage unit 1008.
  • part or all of the computer program may be loaded and/or installed onto the electronic device 1000 via the ROM 1002 and/or the communication unit 1009 .
  • a computer program When a computer program is loaded into RAM 1003 and computed by When unit 1001 is executed, one or more steps of method 200 described above may be performed.
  • computing unit 1001 may be configured to perform method 200 in any other suitable manner (eg, by means of firmware).
  • Various implementations of the systems and techniques described above may be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on a chip implemented in a system (SOC), complex programmable logic device (CPLD), computer hardware, firmware, software, and/or combinations thereof.
  • FPGAs field programmable gate arrays
  • ASICs application specific integrated circuits
  • ASSPs application specific standard products
  • SOC system
  • CPLD complex programmable logic device
  • computer hardware firmware, software, and/or combinations thereof.
  • These various embodiments may include implementation in one or more computer programs executable and/or interpreted on a programmable system including at least one programmable processor, the programmable processor
  • the processor which may be a special purpose or general purpose programmable processor, may receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.
  • An output device may be a special purpose or general purpose programmable processor, may receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.
  • An output device may be a special purpose or general purpose programmable processor, may receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.
  • Program code for implementing the methods of the present disclosure may be written in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing device, such that the program codes, when executed by the processor or controller, cause the functions specified in the flowcharts and/or block diagrams/ The operation is implemented.
  • the program code may execute entirely on the machine, partly on the machine, as a stand-alone software package, partly on the machine and partly on a remote machine or entirely on the remote machine or server.
  • a machine-readable medium may be a tangible medium that may contain or store a program for use by or in connection with an instruction execution system, apparatus, or device.
  • the machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium.
  • Machine-readable media may include, but are not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or devices, or any suitable combination of the foregoing.
  • machine-readable storage media would include one or more wire-based electrical connections, laptop disks, hard drives, random access memory (RAM), read only memory (ROM), erasable programmable read only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination of the above.
  • RAM random access memory
  • ROM read only memory
  • EPROM or flash memory erasable programmable read only memory
  • CD-ROM portable compact disk read-only memory
  • magnetic storage device or any suitable combination of the above.
  • the systems and techniques described herein may be implemented on a computer having a display device (eg, a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user ); and keyboard and pointing devices (e.g., mouse or Trackball), the user can provide input to the computer through the keyboard and the pointing device.
  • a display device eg, a CRT (cathode ray tube) or LCD (liquid crystal display) monitor
  • keyboard and pointing devices e.g., mouse or Trackball
  • Other kinds of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and may be provided in any form, including Acoustic input, voice input or tactile input) to receive input from the user.
  • the systems and techniques described herein may be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., A user's computer having a graphical user interface or web browser through which the user can interact with implementations of the systems and technologies described herein), or including such backend components, middleware components, or any combination of front-end components in a computing system.
  • the components of the system may be interconnected by any form or medium of digital data communication (eg, a communications network). Examples of communication networks include: local area network (LAN), wide area network (WAN), and the Internet.
  • Computer systems may include clients and servers.
  • Clients and servers are generally remote from each other and typically interact over a communications network.
  • the relationship of client and server is created by computer programs running on corresponding computers and having a client-server relationship with each other.
  • the server can be a cloud server, a distributed system server, or a server combined with a blockchain.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • Computational Linguistics (AREA)
  • Databases & Information Systems (AREA)
  • Artificial Intelligence (AREA)
  • Audiology, Speech & Language Pathology (AREA)
  • General Health & Medical Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Business, Economics & Management (AREA)
  • Finance (AREA)
  • Data Mining & Analysis (AREA)
  • Accounting & Taxation (AREA)
  • Probability & Statistics with Applications (AREA)
  • Development Economics (AREA)
  • Economics (AREA)
  • Marketing (AREA)
  • Strategic Management (AREA)
  • General Business, Economics & Management (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

本公开提供了一种对象推荐方法、装置、电子设备和存储介质，涉及人工智能领域，尤其涉及基于人工智能的推荐技术领域。实现方案为：获得待推荐的对象，对象具有相应的描述文本；获得描述文本所包括的多个描述分词中的多个目标描述分词，多个目标描述分词用于将描述文本与其他对象对应的其他描述文本进行区分；基于多个目标描述分词，获得对象的对象表征向量；以及基于对象表征向量，推荐对象。

Description

对象推荐方法、装置、电子设备和存储介质
相关申请的交叉引用
本申请要求于2022年8月3日提交的中国专利申请202210927036.9的优先权，其全部内容通过引用整体结合在本申请中。
技术领域
本公开涉及人工智能技术领域，尤其涉及基于人工智能的推荐技术领域，具体涉及一种对象推荐方法、装置、电子设备、计算机可读存储介质和计算机程序产品。
背景技术
人工智能是研究使计算机来模拟人的某些思维过程和智能行为(如学习、推理、思考、规划等)的学科，既有硬件层面的技术也有软件层面的技术。人工智能硬件技术一般包括如传感器、专用人工智能芯片、云计算、分布式存储、大数据处理等技术；人工智能软件技术主要包括计算机视觉技术、语音识别技术、自然语言处理技术以及机器学习/深度学习、大数据处理技术、知识图谱技术等几大方向。
基于人工智能的推荐技术，已经渗透到各个领域。其中，基于人工智能的对象推荐，通过预测用户对对象的偏好，向用户推荐符合其偏好的对象。
在此部分中描述的方法不一定是之前已经设想到或采用的方法。除非另有指明，否则不应假定此部分中描述的任何方法仅因其包括在此部分中就被认为是现有技术。类似地，除非另有指明，否则此部分中提及的问题不应认为在任何现有技术中已被公认。
发明内容
本公开提供了一种对象推荐方法、装置、电子设备、计算机可读存储介质和计算机程序产品。
根据本公开的一方面，提供了一种对象推荐方法，包括：获得待推荐的对象，所述对象具有相应的描述文本；获得所述描述文本所包括的多个描述分词中的多个目标描述分词，所述多个目标描述分词用于将所述描述文本与其他对象对应的其他描述文本进行区分；基于所述多个目标描述分词，获得所述对象的对象表征向量；以及基于所述对象表征向量，推荐所述对象。
根据本公开的另一方面，提供了一种对象推荐装置，包括：对象获取单元，被配置用于获得待推荐的对象，所述对象具有相应的描述文本；目标描述分词获取单元，被配置用于获得所述描述文本所包括的多个描述分词中的多个目标描述分词，所述多个目标描述分词用于将所述描述文本与其他对象对应的其他描述文本进行区分；对象表征向量获取单元，被配置用于基于所述多个目标描述分词，获得所述对象的对象表征向量；以及推荐单元，被配置用于基于所述对象表征向量，推荐所述对象。
根据本公开的另一方面，提供了一种电子设备，包括：至少一个处理器；以及与所述至少一个处理器通信连接的存储器；其中所述存储器存储有可被所述至少一个处理器执行的指令，所述指令被所述至少一个处理器执行，以使所述至少一个处理器能够执行根据本公开的实施例所述的方法。
根据本公开的另一方面，提供了一种存储有计算机指令的非瞬时计算机可读存储介质，其中，所述计算机指令用于使所述计算机执行根据本公开的实施例所述的方法。
根据本公开的另一方面，提供了一种计算机程序产品，包括计算机程序，其中，所述计算机程序在被处理器执行时实现根据本公开的实施例所述的方法。
根据本公开的一个或多个实施例，可以提升为用户所推荐的对象的准确性。
应当理解，本部分所描述的内容并非旨在标识本公开的实施例的关键或重要特征，也不用于限制本公开的范围。本公开的其它特征将通过以下的说明书而变得容易理解。
附图说明
附图示例性地示出了实施例并且构成说明书的一部分，与说明书的文字描述一起用于讲解实施例的示例性实施方式。所示出的实施例仅出于例示的目的，并不限制权利要求的范围。在所有附图中，相同的附图标记指代类似但不一定相同的要素。
图1示出了根据本公开的实施例的可以在其中实施本文描述的各种方法的示例性***的示意图；
图2示出了根据本公开的实施例的对象推荐方法的流程图；
图3示出了根据本公开的实施例的对象推荐方法中获得描述文本所包括的多个描述分词中的多个目标描述分词的过程的流程图；
图4示出了根据本公开的实施例的对象推荐方法中基于多个目标描述分词，获得对象的对象表征向量的过程的流程图；
图5示出了根据本公开的实施例的对象推荐方法中基于多个目标描述分词中的每一个目标描述分词的语义表示向量，获得对象表征向量的过程的流程图；
图6示出了根据本公开的实施例的对象推荐方法的流程图；
图7示出了根据本公开的实施例的对象推荐方法中获得搜索文本所包括的多个搜索分词中的多个目标搜索分词的过程的流程图；
图8示出了根据本公开的实施例的对象推荐方法中基于多个目标搜索分词，获得搜索文本的搜索文本表征向量的过程的流程图；
图9示出了根据本公开的实施例的对象推荐装置的结构框图；以及
图10示出了能够用于实现本公开的实施例的示例性电子设备的结构框图。
具体实施方式
以下结合附图对本公开的示范性实施例做出说明，其中包括本公开实施例的各种细节以助于理解，应当将它们认为仅仅是示范性的。因此，本领域普通技术人员应当认识到，可以对这里描述的实施例做出各种改变和修改，而不会背离本公开的范围。同样，为了清楚和简明，以下的描述中省略了对公知功能和结构的描述。
在本公开中，除非另有说明，否则使用术语“第一”、“第二”等来描述各种要素不意图限定这些要素的位置关系、时序关系或重要性关系，这种术语只是用于将一个要素与另一要素区分开。在一些示例中，第一要素和第二要素可以指向该要素的同一实例，而在某些情况下，基于上下文的描述，它们也可以指代不同实例。
在本公开中对各种所述示例的描述中所使用的术语只是为了描述特定示例的目的，而并非旨在进行限制。除非上下文另外明确地表明，如果不特意限定要素的数量，则该要素可以是一个也可以是多个。此外，本公开中所使用的术语“和/或”涵盖所列出的项目中的任何一个以及全部可能的组合方式。
下面将结合附图详细描述本公开的实施例。
图1示出了根据本公开的实施例可以将本文描述的各种方法和装置在其中实施的示例性***100的示意图。参考图1，该***100包括一个或多个客户端设备101、102、103、104、105和106、服务器120以及将一个或多个客户端设备耦接到服务器120的一个或多个通信网络110。客户端设备101、102、103、104、105和106可以被配置为执行一个或多个应用程序。
在本公开的实施例中，服务器120可以运行使得能够执行根据本公开的对象推荐方法的一个或多个服务或软件应用。
在某些实施例中，服务器120还可以提供其他服务或软件应用，这些服务或软件应用可以包括非虚拟环境和虚拟环境。在某些实施例中，这些服务可以作为基于web的服务或云服务提供，例如在软件即服务(SaaS)模型下提供给客户端设备101、102、103、104、105和/或106的用户。
在图1所示的配置中，服务器120可以包括实现由服务器120执行的功能的一个或多个组件。这些组件可以包括可由一个或多个处理器执行的软件组件、硬件组件或其组合。操作客户端设备101、102、103、104、105和/或106的用户可以依次利用一个或多个客户端应用程序来与服务器120进行交互以利用这些组件提供的服务。应当理解，各种不同的***配置是可能的，其可以与***100不同。因此，图1是用于实施本文所描述的各种方法的***的一个示例，并且不旨在进行限制。
用户可以使用客户端设备101、102、103、104、105和/或106来接收向该用户推荐的一个或多个对象。客户端设备可以提供使客户端设备的用户能够与客户端设备进行交互的接口。客户端设备还可以经由该接口向用户输出信息。尽管图1仅描绘了六种客户端设备，但是本领域技术人员将能够理解，本公开可以支持任何数量的客户端设备。
客户端设备101、102、103、104、105和/或106可以包括各种类型的计算机设备，例如便携式手持设备、通用计算机(诸如个人计算机和膝上型计算机)、工作站计算机、可穿戴设备、智能屏设备、自助服务终端设备、服务机器人、游戏***、瘦客户端、各种消息收发设备、传感器或其他感测设备等。这些计算机设备可以运行各种类型和版本的软件应用程序和操作***，例如MICROSOFT Windows、APPLE iOS、类UNIX操作***、Linux或类Linux操作***(例如GOOGLE Chrome OS)；或包括各种移动操作***，例如MICROSOFT Windows Mobile OS、iOS、Windows Phone、Android。便携式手持设备可以包括蜂窝电话、智能电话、平板电脑、个人数字助理(PDA)等。可穿戴设备可以包括头戴式显示器(诸如智能眼镜)和其他设备。游戏***可以包括各种手持式游戏设备、支持互联网的游戏设备等。客户端设备能够执行各种不同的应用程序，例如各种与Internet相关的应用程序、通信应用程序(例如电子邮件应用程序)、短消息服务(SMS)应用程序，并且可以使用各种通信协议。
网络110可以是本领域技术人员熟知的任何类型的网络，其可以使用多种可用协议中的任何一种(包括但不限于TCP/IP、SNA、IPX等)来支持数据通信。仅作为示例，一个或多个网络110可以是局域网(LAN)、基于以太网的网络、令牌环、广域网(WAN)、因特网、虚拟网络、虚拟专用网络(VPN)、内部网、外部网、区块链网络、公共交换电话网(PSTN)、红外网络、无线网络(例如蓝牙、WIFI)和/或这些和/或其他网络的任意组合。
服务器120可以包括一个或多个通用计算机、专用服务器计算机(例如PC(个人计算机)服务器、UNIX服务器、中端服务器)、刀片式服务器、大型计算机、服务器群集或任何其他适当的布置和/或组合。服务器120可以包括运行虚拟操作***的一个或多个虚拟机，或者涉及虚拟化的其他计算架构(例如可以被虚拟化以维护服务器的虚拟存储设备的逻辑存储设备的一个 或多个灵活池)。在各种实施例中，服务器120可以运行提供下文所描述的功能的一个或多个服务或软件应用。
服务器120中的计算单元可以运行包括上述任何操作***以及任何商业上可用的服务器操作***的一个或多个操作***。服务器120还可以运行各种附加服务器应用程序和/或中间层应用程序中的任何一个，包括HTTP服务器、FTP服务器、CGI服务器、JAVA服务器、数据库服务器等。
在一些实施方式中，服务器120可以包括一个或多个应用程序，以分析和合并从客户端设备101、102、103、104、105和/或106的用户接收的数据馈送和/或事件更新。服务器120还可以包括一个或多个应用程序，以经由客户端设备101、102、103、104、105和/或106的一个或多个显示设备来显示数据馈送和/或实时事件。
在一些实施方式中，服务器120可以为分布式***的服务器，或者是结合了区块链的服务器。服务器120也可以是云服务器，或者是带人工智能技术的智能云计算服务器或智能云主机。云服务器是云计算服务体系中的一项主机产品，以解决传统物理主机与虚拟专用服务器(VPS，Virtual Private Server)服务中存在的管理难度大、业务扩展性弱的缺陷。
***100还可以包括一个或多个数据库130。在某些实施例中，这些数据库可以用于存储数据和其他信息。例如，数据库130中的一个或多个可用于存储诸如音频文件和视频文件的信息。数据库130可以驻留在各种位置。例如，由服务器120使用的数据库可以在服务器120本地，或者可以远离服务器120且可以经由基于网络或专用的连接与服务器120通信。数据库130可以是不同的类型。在某些实施例中，由服务器120使用的数据库例如可以是关系数据库。这些数据库中的一个或多个可以响应于命令而存储、更新和检索到数据库以及来自数据库的数据。
在某些实施例中，数据库130中的一个或多个还可以由应用程序使用来存储应用程序数据。由应用程序使用的数据库可以是不同类型的数据库，例如键值存储库，对象存储库或由文件***支持的常规存储库。
图1的***100可以以各种方式配置和操作，以使得能够应用根据本公开所描述的各种方法和装置。
根据本公开的一方面，提供了一种对象推荐方法。参看图2，根据本公开的一些实施例的对象推荐方法200包括：
步骤S210：获得待推荐的对象，所述对象具有相应的描述文本；
步骤S220：获得所述描述文本所包括的多个描述分词中的多个目标描述分词，所述多个目标描述分词用于将所述描述文本与其他对象对应的其他描述文本进行区分；
步骤S230：基于所述多个目标描述分词，获得所述对象的对象表征向量；以及
步骤S240：基于所述对象表征向量，推荐所述对象。
在相关技术中，往往基于描述文本获得对象的表征向量，并基于对象的表征向量，确定是否将对象推荐给用户。由于描述文本中常常包含与该对象无关的无法将该对象与其他对象进行区分的词，例如，描述对象的应用场景的词，使得所获得的对象的表征向量不具有代表性；基于该表征向量获得为用户推荐的对象的过程中，往往使得为用户推荐的对象与其预期相差较远，尤其是在通过将该表征向量与用户的搜索请求进行匹配的过程中，由于用户的搜索请求往往较为简练，往往匹配不到与其精确匹配的表征向量，使得为用户推荐的对象不够准确。
根据本公开的实施例中，通过获得对象的描述文本中的多个描述分词中的多个目标描述分词，该多个目标描述分词能够将对象的描述文本与其他对象的其他描述文本区分开来，并基于该多个目标描述分词，获得对象的对象表征向量，使得所获得的对象表征向量能够将对象与其他对象进行区分，实现对对象进行有重点的表征，从而基于该对象表征向量进行推荐的过程中，能够进行精准匹配，提高向用户推荐的对象的准确性。
在一些实施例中，待推荐的对象可以是任何以电子数据形式存在的任何信息、资源等，例如，视频、文章、商品等。通过将待推荐的对象通过网络传送给各个客户端，实现对待推荐的对象的推荐。
在一些实施例中，当对象为视频时，对象的描述文本可以是视频的标题、字幕、标签等；当对象为商品时，描述文本可以是商品的标题；当对象为违章时，描述文本可以是文章的标题或者文章的内容文本。
在根据本公开的一个示例中，所述对象包括商品，所述描述文本包括所述商品的标题。
在一些实施例中，如图3所示，获得所述描述文本所包括的多个描述分词中的多个目标描述分词包括：
步骤S310：对所述描述文本进行切词，以获得所述多个描述分词；
步骤S320：获得所述多个描述分词中的每一个描述分词对应的第一得分，所述第一得分指示基于相应描述分词将所述描述文本与其他对象对应的其他描述文本进行区分的可能性；以及
步骤S330：基于所述多个描述分词对应的多个第一得分，获得所述多个目标描述分词。
通过获得描述文本中各个描述分词的第一得分，该第一得分指示基于相应描述分词将所述描述文本与其他对象对应的其他描述文本进行区分的可能性，并且基于各个描述分词的第一得分获得多个目标描述分词，使获得的多个目标描述分词准确。
在一些实施例中，将各个描述分词输入至经训练的打分模型，以获得每一个描述分词对应的第一得分。
在一些实施例中，通过词频-逆向文件频率(TF-IDF)获得多个描述分词中的每一个描述分词对应的第一得分。
例如，通过获得对象集合中的每一个对象的描述文本，以获得描述文本集合，基于该描述文本集合，通过公式(1)获得第i个描述文本I中的描述分词x的第一得分S：
其中，xi为描述分词x在描述文本I中出现的次数，In为描述文本I中的各个描述分词出现的次数总和，N为描述文本集合中的文本总数，Nx为包括描述分词x的描述文本的总数。
通过上述统计方法，可以获得描述文本中每个描述分词对于将该描述文本和其他描述文本进行分所占的权重(即第一得分S)，权重越高表明基于该描述分词将该描述文本与其他描述文本进行区分的可能性越大。
在一些实施例中，将多个描述分词中第一得分大于预设得分阈值的描述分词确定为目标描述分词，以获得所述多个目标描述分词。
在一些实施例中，所述多个目标描述分词包括所述多个描述分词中预设数量的多个描述分词，所述预设数量的描述分词中的每一个描述分词的第一得分高于所述多个描述分词中的其他描述分词，所述其他描述分词区别于所述预设数量的描述分词中的每一个描述分词。
多个目标描述分词是多个描述分词中第一得分较大的预设数量的多个描述分词，使获得的目标描述分词是多个描述分词中更具代表性的描述分词，进而使基于多个目标描述分词所获得的对象的对象表征向量更加准确。
在一些实施例中，直接基于多个目标描述分词中的每一个目标描述分词的词向量，获得对象的对象表征向量。
在一些实施例中，如图4所示，基于所述多个目标描述分词，获得所述对象的对象表征向量包括：
步骤S410：获得所述多个目标描述分词中的每一个目标描述分词的语义表示向量，该语义表示向量与该目标描述分词在所述描述文本中的位置相关；以及
步骤S420：基于所述多个目标描述分词中的每一个目标描述分词的语义表示向量，获得所述对象表征向量。
通过获得每一个目标描述分词的语义表示向量，语义表示向量与该目标描述分词在所述描述文本中的位置相关，使得每一个目标描述分词的语义表示向量与该描述文本的语义相关，能够挖掘描述文本中深层语义之间的相似信息，进而使基于每一个基于目标描述分词的语义表示向量获得的对象表征向量包含描述文本中的深层语义之间的相似信息，使对象表征向量对对象的表征准确。
在一些实施例中，基于BERT深度学习模型，获得多个目标描述分词中的每一个目标描述分词的语义向量表示。
例如，通过将描述文本进行切词，获得多个描述分词之后，将多个描述分词按照其在描述文本中的顺序组成的词序列，并输入BERT模型，从而获得每一个描述分词的语义表征向量，其中，包括多个目标描述分词中的每一个目标描述分词的语义表征向量。
在一些实施例中，获得多个目标描述分词中的每一个目标描述分词的语义表征向量后，通过将多个目标描述分词的多个语义表征向量直接相加，获得对象的对象表征向量。
在一些实施例中，基于所述多个目标描述分词中的每一个目标描述分词的语义表示向量和第一得分，获得对象表征向量。
在一些实施例中，如图5所示，基于所述多个目标描述分词中的每一个目标描述分词的语义表示向量，获得所述对象表征向量包括：
步骤S510：对所述多个目标描述分词中的每一个目标描述分词的第一得分进行归一化处理，以获得该描述分词的加权得分；
步骤S520：基于所述多个目标描述分词中的每一个目标描述分词的加权得分对该目标描述分词的语义对象表征向量进行加权处理，以获得该目标描述分词的加权向量；以及
步骤S530：基于所述多个目标描述分词中的每一个目标描述分词的加权向量，获得所述对象表征向量。
通过基于对每一个目标描述分词的第一得分进行归一化处理之后获得的加权得分，对相应的语义表示向量进行加权处理，并基于加权处理后获得的加权向量获得对象表征向量，使获得的对象表征向量还包括各个目标描述分词之间对描述文本的重要性程度(基于该目标描述分词将该描述文本与其他描述文本进行区分的可能性)的区分，进一步提升所获得的对象表征向量的准确性。
在一些实施例中，通过直接将多个目标描述分词对应的多个加权向量直接相加，获得对象表征向量。
在一些实施例中，在获得对象表征向量之后，还获得用户的用户表征向量，并计算对象表征向量和用户表征向量之间的相似度，基于相似度确定是否将该对象推荐给用户。
在一些实施例中，基于用户的搜索请求，确定是否将该对象推荐给用户。
在一些实施例中，如图6所示，根据本公开的一些实施例的对象推荐方法还包括：
步骤S610：获得用户的搜索文本；
步骤S620：获得所述搜索文本所包括的多个搜索分词中的多个目标搜索分词，所述多个目标搜索分词用于将所述搜索文本与其他搜索文本进行区分；以及
步骤S630：基于所述多个目标搜索分词，获得所述搜索文本的搜索文本表征向量；并且其中，所述基于所述对象表征向量，推荐所述对象包括：
基于所述搜索文本表征向量和所述对象表征向量，确定是否向所述用户推荐所述对象。
根据本公开的实施例，在基于用户的搜索文本为用户推荐对象的过程中，实现文本和文本之间的匹配，由于对于对象的描述文本和用户的搜索文本，分别获得该文本中的多个分词中能够将其与其他相应文本进行区分的多个目标分词，并基于相应的多个目标分词获得相应的表征向量，其中，基于描述文本的多个目标描述分词获得对象表征向量，基于搜索文本的多个目标搜索分词获得搜索文本表征向量，提升对对象和搜索文本进行表征的准确性，使基于搜索文本和对象的描述文本进行匹配时，匹配结果更加准确，进而能够提升为用户推荐的对象的准确性。
在一些实施例中，如图7所示，获得所述搜索文本所包括的多个搜索分词中的多个目标搜索分词包括：
步骤S710：对所述搜索文本进行切词，以获得所述多个搜索分词；
步骤S720：获得所述多个搜索分词中的每一个搜索分词对应的第二得分，所述第二得分指示基于相应搜索分词将所述搜索文本与其他搜索文本进行区分的可能性；以及
步骤S730：基于所述多个搜索分词对应的多个第二得分，获得所述多个目标搜索分词。
通过获得搜索文本中每一个搜索分词的第二得分，该第二得分指示基于相应搜索分词将搜索文本与其他搜索文本进行区分的可能性，并且基于各个搜索分词的第二得分获得多个目标搜索分词，使获得的多个目标搜索分词准确。
在一些实施例中，将各个搜索分词输入至经训练的打分模型，以获得每一个搜索分词对应的第二得分。
在一些实施例中，通过词频-逆向文件频率(TF-IDF)获得多个搜索分词中的每一个搜索分词对应的第二得分。
在一些实施例中，如图8所示，基于所述多个目标搜索分词，获得所述搜索文本的搜索文本表征向量包括：
步骤S810：获得所述多个目标搜索分词中的每一个目标搜索分词的语义表示向量，该语义表示向量与该目标搜索分词在所述搜索文本中的位置相关；
步骤S820：基于所述多个目标搜索分词中的每一个目标搜索分词对应的第二得分，获得该目标搜索分词对应的加权得分；以及
步骤S830：基于所述多个目标搜索分词中的每一个目标搜索分词的语义表示向量和加权得分，获得所述搜索文本表征向量。
通过获得每一个目标搜索分词的语义表示向量，语义表示向量与该目标搜索分词在搜索文本中的位置相关，使得每一个目标搜索分词的语义表示向量与该搜索文本的语义相关，能够挖掘搜索文本中深层语义之间的相似信息，进而使基于每一个基于目标搜索分词的语义表示向量获得的搜索文本表征向量包含搜索文本中的深层语义之间的相似信息，使搜索文本表征向量对搜索文本的表征准确。
同时，通过基于对每一个目标搜索分词的第二得分进行归一化处理之后获得的加权得分，对相应的语义表示向量进行加权处理，并基于加权处理后获得的加权向量获得搜索文本表征向量，使获得的搜索文本表征向量还包括各个目标搜索分词之间对搜索文本的重要性程度(基于该目标搜索分词将该搜索文本与其他搜索文本进行区分的可能性)的区分，进一步提升所获得的搜索文本表征向量的准确性。
在一些实施例中，通过将每一个目标搜索分词的语义表示向量和加权得分相乘，获得加权向量；并将多个目标搜索分词对应的多个加权向量相加，获得搜索文本表征向量。
在一些实施例中，获得搜索文本表征向量之后，计算搜索文本表征向量和对象表征向量之间的相似度，并基于相似度确定是否将对象推荐给用户。
根据本公开的另一方面，还提供一种对象推荐装置，如图9所示，装置900包括：对象获取单元910，被配置用于获得待推荐的对象，所述对象具 有相应的描述文本；目标描述分词获取单元920，被配置用于获得所述描述文本所包括的多个描述分词中的多个目标描述分词，所述多个目标描述分词用于将所述描述文本与其他对象对应的其他描述文本进行区分；对象表征向量获取单元930，被配置用于基于所述多个目标描述分词，获得所述对象的对象表征向量；以及推荐单元940，被配置用于基于所述对象表征向量，推荐所述对象。
在一些实施例中，所述目标描述分词获取单元920包括：切词单元，被配置用于对所述描述文本进行切词，以获得所述多个描述分词；第一得分计算单元，被配置用于获得所述多个描述分词中的每一个描述分词对应的第一得分，所述第一得分指示基于相应描述分词将所述描述文本与其他对象对应的其他描述文本进行区分的可能性；以及目标描述分词获取子单元，被配置用于基于所述多个描述分词对应的多个第一得分，获得所述多个目标描述分词。
在一些实施例中，所述多个目标描述分词包括所述多个描述分词中预设数量的多个描述分词，所述预设数量的描述分词中的每一个描述分词的第一得分高于所述多个描述分词中的其他描述分词，所述其他描述分词区别于所述预设数量的描述分词中的每一个描述分词。
在一些实施例中，所述对象表征向量获取单元930包括：语义表示向量获取单元，被配置用于获得所述多个目标描述分词中的每一个目标描述分词的语义表示向量，该语义表示向量与该目标描述分词在所述描述文本中的位置相关；以及对象表征向量获取子单元，被配置用于基于所述多个目标描述分词中的每一个目标描述分词的语义表示向量，获得所述对象表征向量。
在一些实施例中，所述对象表征向量获取子单元包括：归一化单元，被配置用于对所述多个目标描述分词中的每一个目标描述分词的第一得分进行归一化处理，以获得该描述分词的加权得分；加权单元，被配置用于基于所述多个目标描述分词中的每一个目标描述分词的加权得分对该目标描述分词的语义对象表征向量进行加权处理，以获得该目标描述分词的加权向量；以及第一获取子单元，被配置用于基于所述多个目标描述分词中的每一个目标描述分词的加权向量，获得所述对象表征向量。
在一些实施例中，还包括：搜索文本获取单元，被配置用于获得用户的搜索文本；目标搜索分词获取单元，被配置用于获得所述搜索文本所包括的多个搜索分词中的多个目标搜索分词，所述多个目标搜索分词用于将所述搜索文本与其他搜索文本进行区分；以及搜索文本表征向量获取单元，被配置用于基于所述多个目标搜索分词，获得所述搜索文本的搜索文本表征向量；并且其中，所述推荐单元930包括：确定单元，被配置用于基于所述搜索文本表征向量和所述对象表征向量，确定是否向所述用户推荐所述对象。
在一些实施例中，所述获得所述搜索文本所包括的多个搜索分词中的多个目标搜索分词包括：对所述搜索文本进行切词，以获得所述多个搜索分词；获得所述多个搜索分词中的每一个搜索分词对应的第二得分，所述第二得分指示基于相应搜索分词将所述搜索文本与其他搜索文本进行区分的可能性；以及基于所述多个搜索分词对应的多个第二得分，获得所述多个目标搜索分词。
在一些实施例中，所述基于所述多个目标搜索分词，获得所述搜索文本的搜索文本表征向量包括：获得所述多个目标搜索分词中的每一个目标搜索分词的语义表示向量，该语义表示向量与该目标搜索分词在所述搜索文本中的位置相关；基于所述多个目标搜索分词中的每一个目标搜索分词对应的第二得分，获得该目标搜索分词对应的加权得分；以及基于所述多个目标搜索分词中的每一个目标搜索分词的语义表示向量和加权得分，获得所述搜索文本表征向量。
本公开的技术方案中，所涉及的用户个人信息的收集、存储、使用、加工、传输、提供和公开等处理，均符合相关法律法规的规定，且不违背公序良俗。
根据本公开的实施例，还提供了一种电子设备、一种可读存储介质和一种计算机程序产品。
参考图10，现将描述可以作为本公开的服务器或客户端的电子设备1000的结构框图，其是可以应用于本公开的各方面的硬件设备的示例。电子设备旨在表示各种形式的数字电子的计算机设备，诸如，膝上型计算机、台式计算机、工作台、个人数字助理、服务器、刀片式服务器、大型计算机、和其它适合的计算机。电子设备还可以表示各种形式的移动装置，诸如，个人数 字处理、蜂窝电话、智能电话、可穿戴设备和其它类似的计算装置。本文所示的部件、它们的连接和关系、以及它们的功能仅仅作为示例，并且不意在限制本文中描述的和/或者要求的本公开的实现。
如图10所示，电子设备1000包括计算单元1001，其可以根据存储在只读存储器(ROM)1002中的计算机程序或者从存储单元1008加载到随机访问存储器(RAM)1003中的计算机程序，来执行各种适当的动作和处理。在RAM 1003中，还可存储电子设备1000操作所需的各种程序和数据。计算单元1001、ROM 1002以及RAM 1003通过总线1004彼此相连。输入/输出(I/O)接口1005也连接至总线1004。
电子设备1000中的多个部件连接至I/O接口1005，包括：输入单元1006、输出单元1007、存储单元1008以及通信单元1009。输入单元1006可以是能向电子设备1000输入信息的任何类型的设备，输入单元1006可以接收输入的数字或字符信息，以及产生与电子设备的用户设置和/或功能控制有关的键信号输入，并且可以包括但不限于鼠标、键盘、触摸屏、轨迹板、轨迹球、操作杆、麦克风和/或遥控器。输出单元1007可以是能呈现信息的任何类型的设备，并且可以包括但不限于显示器、扬声器、视频/音频输出终端、振动器和/或打印机。存储单元1008可以包括但不限于磁盘、光盘。通信单元1009允许电子设备1000通过诸如因特网的计算机网络和/或各种电信网络与其他设备交换信息/数据，并且可以包括但不限于调制解调器、网卡、红外通信设备、无线通信收发机和/或芯片组，例如蓝牙TM设备、802.11设备、WiFi设备、WiMax设备、蜂窝通信设备和/或类似物。
计算单元1001可以是各种具有处理和计算能力的通用和/或专用处理组件。计算单元1001的一些示例包括但不限于中央处理单元(CPU)、图形处理单元(GPU)、各种专用的人工智能(AI)计算芯片、各种运行机器学习模型算法的计算单元、数字信号处理器(DSP)、以及任何适当的处理器、控制器、微控制器等。计算单元1001执行上文所描述的各个方法和处理，例如方法200。例如，在一些实施例中，方法200可被实现为计算机软件程序，其被有形地包含于机器可读介质，例如存储单元1008。在一些实施例中，计算机程序的部分或者全部可以经由ROM 1002和/或通信单元1009而被载入和/或安装到电子设备1000上。当计算机程序加载到RAM 1003并由计算 单元1001执行时，可以执行上文描述的方法200的一个或多个步骤。备选地，在其他实施例中，计算单元1001可以通过其他任何适当的方式(例如，借助于固件)而被配置为执行方法200。
本文中以上描述的***和技术的各种实施方式可以在数字电子电路***、集成电路***、场可编程门阵列(FPGA)、专用集成电路(ASIC)、专用标准产品(ASSP)、芯片上***的***(SOC)、复杂可编程逻辑设备(CPLD)、计算机硬件、固件、软件、和/或它们的组合中实现。这些各种实施方式可以包括：实施在一个或者多个计算机程序中，该一个或者多个计算机程序可在包括至少一个可编程处理器的可编程***上执行和/或解释，该可编程处理器可以是专用或者通用可编程处理器，可以从存储***、至少一个输入装置、和至少一个输出装置接收数据和指令，并且将数据和指令传输至该存储***、该至少一个输入装置、和该至少一个输出装置。
用于实施本公开的方法的程序代码可以采用一个或多个编程语言的任何组合来编写。这些程序代码可以提供给通用计算机、专用计算机或其他可编程数据处理装置的处理器或控制器，使得程序代码当由处理器或控制器执行时使流程图和/或框图中所规定的功能/操作被实施。程序代码可以完全在机器上执行、部分地在机器上执行，作为独立软件包部分地在机器上执行且部分地在远程机器上执行或完全在远程机器或服务器上执行。
在本公开的上下文中，机器可读介质可以是有形的介质，其可以包含或存储以供指令执行***、装置或设备使用或与指令执行***、装置或设备结合地使用的程序。机器可读介质可以是机器可读信号介质或机器可读储存介质。机器可读介质可以包括但不限于电子的、磁性的、光学的、电磁的、红外的、或半导体***、装置或设备，或者上述内容的任何合适组合。机器可读存储介质的更具体示例会包括基于一个或多个线的电气连接、便携式计算机盘、硬盘、随机存取存储器(RAM)、只读存储器(ROM)、可擦除可编程只读存储器(EPROM或快闪存储器)、光纤、便捷式紧凑盘只读存储器(CD-ROM)、光学储存设备、磁储存设备、或上述内容的任何合适组合。
为了提供与用户的交互，可以在计算机上实施此处描述的***和技术，该计算机具有：用于向用户显示信息的显示装置(例如，CRT(阴极射线管)或者LCD(液晶显示器)监视器)；以及键盘和指向装置(例如，鼠标或者 轨迹球)，用户可以通过该键盘和该指向装置来将输入提供给计算机。其它种类的装置还可以用于提供与用户的交互；例如，提供给用户的反馈可以是任何形式的传感反馈(例如，视觉反馈、听觉反馈、或者触觉反馈)；并且可以用任何形式(包括声输入、语音输入或者、触觉输入)来接收来自用户的输入。
可以将此处描述的***和技术实施在包括后台部件的计算***(例如，作为数据服务器)、或者包括中间件部件的计算***(例如，应用服务器)、或者包括前端部件的计算***(例如，具有图形用户界面或者网络浏览器的用户计算机，用户可以通过该图形用户界面或者该网络浏览器来与此处描述的***和技术的实施方式交互)、或者包括这种后台部件、中间件部件、或者前端部件的任何组合的计算***中。可以通过任何形式或者介质的数字数据通信(例如，通信网络)来将***的部件相互连接。通信网络的示例包括：局域网(LAN)、广域网(WAN)和互联网。
计算机***可以包括客户端和服务器。客户端和服务器一般远离彼此并且通常通过通信网络进行交互。通过在相应的计算机上运行并且彼此具有客户端-服务器关系的计算机程序来产生客户端和服务器的关系。服务器可以是云服务器，也可以为分布式***的服务器，或者是结合了区块链的服务器。
应该理解，可以使用上面所示的各种形式的流程，重新排序、增加或删除步骤。例如，本公开中记载的各步骤可以并行地执行、也可以顺序地或以不同的次序执行，只要能够实现本公开公开的技术方案所期望的结果，本文在此不进行限制。
虽然已经参照附图描述了本公开的实施例或示例，但应理解，上述的方法、***和设备仅仅是示例性的实施例或示例，本发明的范围并不由这些实施例或示例限制，而是仅由授权后的权利要求书及其等同范围来限定。实施例或示例中的各种要素可以被省略或者可由其等同要素替代。此外，可以通过不同于本公开中描述的次序来执行各步骤。进一步地，可以以各种方式组合实施例或示例中的各种要素。重要的是随着技术的演进，在此描述的很多要素可以由本公开之后出现的等同要素进行替换。

Claims (20)

  1. 一种对象推荐方法，包括：
    获得待推荐的对象，所述对象具有相应的描述文本；
    获得所述描述文本所包括的多个描述分词中的多个目标描述分词，所述多个目标描述分词用于将所述描述文本与其他对象对应的其他描述文本进行区分；
    基于所述多个目标描述分词，获得所述对象的对象表征向量；以及
    基于所述对象表征向量，推荐所述对象。
  2. 根据权利要求1所述的方法，其中，所述获得所述描述文本所包括的多个描述分词中的多个目标描述分词包括：
    对所述描述文本进行切词，以获得所述多个描述分词；
    获得所述多个描述分词中的每一个描述分词对应的第一得分，所述第一得分指示基于相应描述分词将所述描述文本与其他对象对应的其他描述文本进行区分的可能性；以及
    基于所述多个描述分词对应的多个第一得分，获得所述多个目标描述分词。
  3. 根据权利要求2所述的方法，其中，所述多个目标描述分词包括所述多个描述分词中预设数量的多个描述分词，所述预设数量的描述分词中的每一个描述分词的第一得分高于所述多个描述分词中的其他描述分词，所述其他描述分词区别于所述预设数量的描述分词中的每一个描述分词。
  4. 根据权利要求2所述的方法，其中，所述基于所述多个目标描述分词，获得所述对象的对象表征向量包括：
    获得所述多个目标描述分词中的每一个目标描述分词的语义表示向量，该语义表示向量与该目标描述分词在所述描述文本中的位置相关；以及
    基于所述多个目标描述分词中的每一个目标描述分词的语义表示向量，获得所述对象表征向量。
  5. 根据权利要求3所述的方法，其中，所述基于所述多个目标描述分词中的每一个目标描述分词的语义表示向量，获得所述对象表征向量包括：
    对所述多个目标描述分词中的每一个目标描述分词的第一得分进行归一化处理，以获得该描述分词的加权得分；
    基于所述多个目标描述分词中的每一个目标描述分词的加权得分对该目标描述分词的语义对象表征向量进行加权处理，以获得该目标描述分词的加权向量；以及
    基于所述多个目标描述分词中的每一个目标描述分词的加权向量，获得所述对象表征向量。
  6. 根据权利要求1-5中任一项所述的方法，其中，所述对象包括商品，所述描述文本包括所述商品的标题。
  7. 根据权利要求1-5中任一项所述的方法，还包括：
    获得用户的搜索文本；
    获得所述搜索文本所包括的多个搜索分词中的多个目标搜索分词，所述多个目标搜索分词用于将所述搜索文本与其他搜索文本进行区分；以及
    基于所述多个目标搜索分词，获得所述搜索文本的搜索文本表征向量；并且其中，所述基于所述对象表征向量，推荐所述对象包括：
    基于所述搜索文本表征向量和所述对象表征向量，确定是否向所述用户推荐所述对象。
  8. 根据权利要求7所述的方法，其中，所述获得所述搜索文本所包括的多个搜索分词中的多个目标搜索分词包括：
    对所述搜索文本进行切词，以获得所述多个搜索分词；
    获得所述多个搜索分词中的每一个搜索分词对应的第二得分，所述第二得分指示基于相应搜索分词将所述搜索文本与其他搜索文本进行区分的可能性；以及
    基于所述多个搜索分词对应的多个第二得分，获得所述多个目标搜索分词。
  9. 根据权利要求8所述的方法，其中，所述基于所述多个目标搜索分词，获得所述搜索文本的搜索文本表征向量包括：
    获得所述多个目标搜索分词中的每一个目标搜索分词的语义表示向量，该语义表示向量与该目标搜索分词在所述搜索文本中的位置相关；
    基于所述多个目标搜索分词中的每一个目标搜索分词对应的第二得分，获得该目标搜索分词对应的加权得分；以及
    基于所述多个目标搜索分词中的每一个目标搜索分词的语义表示向量和加权得分，获得所述搜索文本表征向量。
  10. 一种对象推荐装置，包括：
    对象获取单元，被配置用于获得待推荐的对象，所述对象具有相应的描述文本；
    目标描述分词获取单元，被配置用于获得所述描述文本所包括的多个描述分词中的多个目标描述分词，所述多个目标描述分词用于将所述描述文本与其他对象对应的其他描述文本进行区分；
    对象表征向量获取单元，被配置用于基于所述多个目标描述分词，获得所述对象的对象表征向量；以及
    推荐单元，被配置用于基于所述对象表征向量，推荐所述对象。
  11. 根据权利要求10所述的装置，其中，所述目标描述分词获取单元包括：
    切词单元，被配置用于对所述描述文本进行切词，以获得所述多个描述分词；
    第一得分计算单元，被配置用于获得所述多个描述分词中的每一个描述分词对应的第一得分，所述第一得分指示基于相应描述分词将所述描述文本与其他对象对应的其他描述文本进行区分的可能性；以及
    目标描述分词获取子单元，被配置用于基于所述多个描述分词对应的多个第一得分，获得所述多个目标描述分词。
  12. 根据权利要求11所述的装置，其中，所述多个目标描述分词包括所述多个描述分词中预设数量的多个描述分词，所述预设数量的描述分词中的每一个描述分词的第一得分高于所述多个描述分词中的其他描述分词，所述其他描述分词区别于所述预设数量的描述分词中的每一个描述分词。
  13. 根据权利要求11所述的装置，其中，所述对象表征向量获取单元包括：
    语义表示向量获取单元，被配置用于获得所述多个目标描述分词中的每一个目标描述分词的语义表示向量，该语义表示向量与该目标描述分词在所述描述文本中的位置相关；以及
    对象表征向量获取子单元，被配置用于基于所述多个目标描述分词中的每一个目标描述分词的语义表示向量，获得所述对象表征向量。
  14. 根据权利要求12所述的装置，其中，所述对象表征向量获取子单元包括：
    归一化单元，被配置用于对所述多个目标描述分词中的每一个目标描述分词的第一得分进行归一化处理，以获得该描述分词的加权得分；
    加权单元，被配置用于基于所述多个目标描述分词中的每一个目标描述分词的加权得分对该目标描述分词的语义对象表征向量进行加权处理，以获得该目标描述分词的加权向量；以及
    第一获取子单元，被配置用于基于所述多个目标描述分词中的每一个目标描述分词的加权向量，获得所述对象表征向量。
  15. 根据权利要求10-14中任一项所述的装置，还包括：
    搜索文本获取单元，被配置用于获得用户的搜索文本；
    目标搜索分词获取单元，被配置用于获得所述搜索文本所包括的多个搜索分词中的多个目标搜索分词，所述多个目标搜索分词用于将所述搜索文本与其他搜索文本进行区分；以及
    搜索文本表征向量获取单元，被配置用于基于所述多个目标搜索分词，获得所述搜索文本的搜索文本表征向量；并且其中，所述推荐单元包括：
    确定单元，被配置用于基于所述搜索文本表征向量和所述对象表征向量，确定是否向所述用户推荐所述对象。
  16. 根据权利要求15所述的装置，其中，所述获得所述搜索文本所包括的多个搜索分词中的多个目标搜索分词包括：
    对所述搜索文本进行切词，以获得所述多个搜索分词；
    获得所述多个搜索分词中的每一个搜索分词对应的第二得分，所述第二得分指示基于相应搜索分词将所述搜索文本与其他搜索文本进行区分的可能性；以及
    基于所述多个搜索分词对应的多个第二得分，获得所述多个目标搜索分词。
  17. 根据权利要求16所述的装置，其中，所述基于所述多个目标搜索分词，获得所述搜索文本的搜索文本表征向量包括：
    获得所述多个目标搜索分词中的每一个目标搜索分词的语义表示向量，该语义表示向量与该目标搜索分词在所述搜索文本中的位置相关；
    基于所述多个目标搜索分词中的每一个目标搜索分词对应的第二得分，获得该目标搜索分词对应的加权得分；以及
    基于所述多个目标搜索分词中的每一个目标搜索分词的语义表示向量和加权得分，获得所述搜索文本表征向量。
  18. 一种电子设备，包括：
    至少一个处理器；以及
    与所述至少一个处理器通信连接的存储器；其中
    所述存储器存储有可被所述至少一个处理器执行的指令，所述指令被所述至少一个处理器执行，以使所述至少一个处理器能够执行权利要求1-9中任一项所述的方法。
  19. 一种存储有计算机指令的非瞬时计算机可读存储介质，其中，所述计算机指令用于使所述计算机执行根据权利要求1-9中任一项所述的方法。
  20. 一种计算机程序产品，包括计算机程序，其中，所述计算机程序在被处理器执行时实现权利要求1-9中任一项所述的方法。
PCT/CN2023/075417 2022-08-03 2023-02-10 对象推荐方法、装置、电子设备和存储介质 WO2024027125A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202210927036.9A CN115269989B (zh) 2022-08-03 2022-08-03 对象推荐方法、装置、电子设备和存储介质
CN202210927036.9 2022-08-03

Publications (1)

Publication Number Publication Date
WO2024027125A1 true WO2024027125A1 (zh) 2024-02-08

Family

ID=83748758

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2023/075417 WO2024027125A1 (zh) 2022-08-03 2023-02-10 对象推荐方法、装置、电子设备和存储介质

Country Status (2)

Country Link
CN (1) CN115269989B (zh)
WO (1) WO2024027125A1 (zh)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115269989B (zh) * 2022-08-03 2023-05-05 百度在线网络技术（北京）有限公司 对象推荐方法、装置、电子设备和存储介质

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20120016885A1 (en) * 2010-07-16 2012-01-19 Ibm Corporation Adaptive and personalized tag recommendation
CN104408115A (zh) * 2014-11-25 2015-03-11 三星电子（中国）研发中心 一种电视平台上基于语义链接的异构资源推荐方法和装置
CN110147499A (zh) * 2019-05-21 2019-08-20 智者四海（北京）技术有限公司 打标签方法、推荐方法及记录介质
CN113314207A (zh) * 2021-06-28 2021-08-27 挂号网（杭州）科技有限公司 一种对象推荐方法、装置、存储介质及电子设备
CN113449099A (zh) * 2020-03-25 2021-09-28 瑞典爱立信有限公司 文本分类方法和文本分类设备
CN115269989A (zh) * 2022-08-03 2022-11-01 百度在线网络技术（北京）有限公司 对象推荐方法、装置、电子设备和存储介质

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111046221B (zh) * 2019-12-17 2024-06-07 腾讯科技（深圳）有限公司 歌曲推荐方法、装置、终端设备以及存储介质
CN112100524A (zh) * 2020-09-17 2020-12-18 北京百度网讯科技有限公司 一种信息推荐的方法、装置、设备以及存储介质
CN112347778B (zh) * 2020-11-06 2023-06-20 平安科技（深圳）有限公司 关键词抽取方法、装置、终端设备及存储介质
CN113792131B (zh) * 2021-09-23 2024-02-09 深圳平安智慧医健科技有限公司 一种关键词的提取方法、装置、电子设备及存储介质
CN114461783A (zh) * 2022-01-14 2022-05-10 腾讯科技（深圳）有限公司 关键词生成方法、装置、计算机设备、存储介质和产品

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20120016885A1 (en) * 2010-07-16 2012-01-19 Ibm Corporation Adaptive and personalized tag recommendation
CN104408115A (zh) * 2014-11-25 2015-03-11 三星电子（中国）研发中心 一种电视平台上基于语义链接的异构资源推荐方法和装置
CN110147499A (zh) * 2019-05-21 2019-08-20 智者四海（北京）技术有限公司 打标签方法、推荐方法及记录介质
CN113449099A (zh) * 2020-03-25 2021-09-28 瑞典爱立信有限公司 文本分类方法和文本分类设备
CN113314207A (zh) * 2021-06-28 2021-08-27 挂号网（杭州）科技有限公司 一种对象推荐方法、装置、存储介质及电子设备
CN115269989A (zh) * 2022-08-03 2022-11-01 百度在线网络技术（北京）有限公司 对象推荐方法、装置、电子设备和存储介质

Also Published As

Publication number Publication date
CN115269989A (zh) 2022-11-01
CN115269989B (zh) 2023-05-05

Similar Documents

Publication Publication Date Title
US20230005284A1 (en) Method for training image-text matching model, computing device, and storage medium
WO2023245938A1 (zh) 对象推荐方法和装置
WO2023019948A1 (zh) 多模态信息库的检索方法、管理方法、装置、设备和介质
WO2023142406A1 (zh) 排序方法、排序模型的训练方法、装置、电子设备及介质
US20230052389A1 (en) Human-object interaction detection
CN116501960B (zh) 内容检索方法、装置、设备及介质
US20230051232A1 (en) Human-object interaction detection
US20230047628A1 (en) Human-object interaction detection
CN112836072A (zh) 信息展示方法及装置、电子设备和介质
EP4242879A1 (en) Ranking method and apparatus, ranking model training method and apparatus, and electronic device and medium
CN113190695A (zh) 多媒体数据搜索方法及装置、计算机设备和介质
WO2024027125A1 (zh) 对象推荐方法、装置、电子设备和存储介质
WO2023050732A1 (zh) 对象推荐方法和装置
WO2023240833A1 (zh) 信息推荐方法及装置、电子设备和介质
CN116541536A (zh) 知识增强的内容生成***、数据生成方法、设备和介质
US20220004801A1 (en) Image processing and training for a neural network
CN115578501A (zh) 图像处理方法、装置、电子设备和存储介质
CN112860681B (zh) 数据清洗方法及装置、计算机设备和介质
CN114281990A (zh) 文档分类方法及装置、电子设备和介质
CN115809364B (zh) 对象推荐方法和模型训练方法
US11790014B2 (en) System and method of determining content similarity by comparing semantic entity attributes
CN114861071B (zh) 对象推荐方法和装置
CN113535958B (zh) 生产线索聚合方法、装置及***、电子设备和介质
US20230044508A1 (en) Data labeling processing
CN115203544A (zh) 推荐方法及装置、电子设备和介质

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: 23848840

Country of ref document: EP

Kind code of ref document: A1