CN112785324A

CN112785324A - Information processing method, information processing apparatus, electronic device, and medium

Info

Publication number: CN112785324A
Application number: CN201911083909.7A
Authority: CN
Inventors: 贺松林; 谭伟; 吴鹏; 雷章明
Original assignee: Beijing Wodong Tianjun Information Technology Co Ltd
Current assignee: Beijing Wodong Tianjun Information Technology Co Ltd
Priority date: 2019-11-07
Filing date: 2019-11-07
Publication date: 2021-05-11

Abstract

An information processing method includes receiving an information request including request subject information and transmission object information; in response to the information request, determining information to be shown corresponding to the sending object information, wherein the information to be shown is specific to one or more objects to be recommended, the showing information comprises a template and element information, the template is determined based on the main body information, and the element information is determined based on at least one of the sending object information and the objects to be recommended; and feeding back the information to be displayed to a request main body corresponding to the request main body information.

Description

Information processing method, information processing apparatus, electronic device, and medium

Technical Field

The present disclosure relates to the field of internet technologies, and in particular, to an information processing method, an information processing apparatus, an electronic device, and a medium.

Background

The existing information flow advertisement delivery flow and framework basically take an advertisement position as a basic resource center, and different creatives are created on the advertisement position to meet different media. Then, popularization is carried out according to the marketing plan and the marketing unit.

In implementing the disclosed concept, the inventors found that there are at least the following problems in the related art: marketing plans are often too long to build and have too many settings, such as time, title, description, etc., and therefore require a significant amount of labor and time to repeatedly prepare creatives and materials for different ad slots. In addition, when accurate delivery is needed again, the method of the related art is that the same version of advertisement material is used for delivery, which easily causes aesthetic fatigue, and even if the advertisement material is replaced regularly, the personalized delivery of thousands of people cannot be realized, and the cost for replacing the material is high.

Disclosure of Invention

In view of the above, the present disclosure provides an information processing method, an information processing apparatus, an electronic device, and a medium that can provide a friendly management manner and can provide diversified advertisement materials.

One aspect of the present disclosure provides an information processing method performed by a server side, including: firstly, an information request is received, the information request comprises request main body information and sending object information, and information to be displayed corresponding to the sending object information is determined in response to the information request. The information to be shown is specific to one or more objects to be recommended, the showing information comprises a template and element information, the template is determined based on the main body information, and the element information is determined based on at least one of the sending object information and the objects to be recommended. And then, feeding back the information to be displayed to a request main body corresponding to the request main body information.

According to the embodiment of the disclosure, information such as the advertisement space of the request subject can be input in advance, the advertisement space and the corresponding template can be conveniently determined according to the request subject, and the management convenience is improved. In addition, the object to be recommended and the corresponding material are determined according to the sending object information, and the template and the material are assembled in real time to obtain the advertisement content, so that thousands of people and thousands of faces of advertisement content can be realized, and the user experience is improved.

According to an embodiment of the present disclosure, the method further comprises: before receiving an information request, storing a first mapping relation between request body information and information transmission associated information, wherein the information transmission associated information comprises a template. Correspondingly, the determining the information to be shown corresponding to the sending object information includes: firstly, one or more objects to be recommended for the sending object information are obtained, and if the objects to be recommended are obtained and the number of the obtained objects to be recommended is greater than or equal to a number threshold value, a template is determined according to the first mapping relation and the request main body information. Then, acquiring element information matched with the sending object information and the object to be recommended, and then assembling the template and the matched element information to obtain the information to be displayed for the object to be recommended.

According to an embodiment of the present disclosure, the acquiring one or more objects to be recommended for the transmission object information includes: firstly, an object to be recommended is obtained, and then the object to be recommended is filtered based on a preset rule to obtain a recommended object. Wherein the preset rule comprises at least one of: for example, whether the repeatedly acquired associated information of the object to be recommended meets a repeatedly acquiring condition, and for example, whether the object to be recommended belongs to a recommendation prohibition object, where the recommendation prohibition object includes at least one of: the sending object forbids the recommending object, the terminal model forbids the recommending object, the request main body forbids the recommending object and the policy forbids the recommending object, and for example, whether the inventory information of the object to be recommended meets the inventory condition or not.

According to an embodiment of the present disclosure, the acquiring an object to be recommended includes: the method comprises the steps of firstly, obtaining historical behavior commodity sequence information corresponding to sending object information, and then processing the sending object information and the historical behavior commodity sequence information by using a first model to obtain object information to be recommended.

According to an embodiment of the present disclosure, the first model includes a predictor sub-model and a parameter update sub-model. The input of the predictor model comprises sending object information and historical behavior commodity sequence information, the historical behavior commodity sequence information comprises target marking information, and the output of the predictor model comprises a recommended object and a corresponding accuracy prediction result. The input of the parameter updating submodel comprises an accuracy prediction result and target labeling information corresponding to the sending object information, and the output of the parameter updating submodel comprises an optimization model parameter, wherein the optimization model parameter is obtained based on a first loss function, and the first loss function is constructed based on the accuracy prediction result and the target labeling information corresponding to the sending object information.

According to an embodiment of the present disclosure, the optimization model parameter corresponds to a direction in which a gradient of the first loss function decreases fastest, and corresponds to the first loss function taking a small value.

According to an embodiment of the present disclosure, the repeatedly acquiring the associated information includes a repeated purchase cycle. Accordingly, the method further comprises: and inputting the sending object information and the information of the object to be recommended into a second model so as to obtain a repurchase period corresponding to the information of the object to be recommended.

According to an embodiment of the present disclosure, the method further comprises training the second model by: training the second model with training data such that the output of the second model approaches the labeled information of the training data. The training data comprises sending object information, object information to be recommended, labeling information and at least one of the following: the commodity sequence information of the historical behaviors of the users, the commodity sequence information of the historical behaviors of the users in the same category and the commodity sequence information of the historical behaviors of all the users.

According to an embodiment of the present disclosure, the method further comprises: and if the objects to be recommended are not acquired or the number of the objects to be recommended is less than the number threshold, acquiring the information to be displayed of the hot object, wherein the information to be displayed of the hot object is matched with the information of the sending object. Then, the information to be displayed of the hot object is used as the information to be displayed corresponding to the information of the sending object, or the information to be displayed of the hot object is used for complementing the information to be displayed of the object to be recommended, so that the quantity of the information to be displayed reaches the quantity threshold value.

According to an embodiment of the present disclosure, the method further comprises: before receiving an information request, receiving and storing contract information, then generating to-be-audited scheduling information of a request subject based on the contract information, then receiving an auditing result aiming at the to-be-audited scheduling information, and if the auditing result is passed, taking the to-be-audited scheduling information as scheduling information. Accordingly, the method further comprises: and after receiving an information request, matching the request main body information and the request time with the scheduling information, and responding to the information request if matching is successful.

According to an embodiment of the present disclosure, the method further comprises: after storing a first mapping relation between request subject information and information transmission associated information, receiving an updating operation, and responding to the updating operation to update the first mapping relation.

According to an embodiment of the present disclosure, the method further comprises: after the request main body sends the information to be displayed to the main body sending the object information, user operation information is received, and historical behavior commodity sequence information corresponding to the sending object information is updated in response to the user operation information, so that at least the information to be recommended is updated.

According to an embodiment of the present disclosure, the method further comprises: after receiving an information request, load splitting is performed in response to the information request.

According to an embodiment of the present disclosure, the load splitting includes splitting the load in a common component mode. Each node of the server side is provided with an independent load balancer, and the load balancer supports single-node multi-instance deployment so as to distribute loads of multiple instances under the node.

According to the embodiment of the disclosure, the element information adopts a multi-level cache. The multi-level cache comprises a server-side cache and a local cache, the server-side cache comprises at least two Redis clusters which are respectively used for storing at least part of the element information, and the server-side cache responds to a local request and transmits cached element information corresponding to the local request to the local cache.

Another aspect of the present disclosure provides an information processing method performed by a media side, including: firstly, an information request is sent to the server side, and the information request comprises request main body information and sending object information. Then, receiving information to be shown for the sending object from a server, wherein the information to be shown is for one or more objects to be recommended, the showing information comprises a template and element information, the template is determined based on the subject information, and the element information is determined based on at least one of the sending object information and the objects to be recommended. And then, sending the information to be displayed to the client side of the sending object so as to facilitate the client side of the sending object to display the information to be displayed.

According to an embodiment of the present disclosure, the method further comprises: after the information to be displayed is sent to the client side of the sending object, user operation information is received and sent to the server side, so that the server side can update the historical behavior commodity sequence information corresponding to the information of the sending object and the information of the object to be recommended.

Another aspect of the present disclosure provides an information processing apparatus including: the display device comprises an information request receiving module, an information to be displayed determining module and a first sending module. The information request receiving module is used for receiving an information request, wherein the information request comprises request main body information and sending object information. The information to be shown determining module is configured to determine information to be shown corresponding to the sending object information in response to the information request, where the information to be shown is for one or more objects to be recommended, the display information includes a template and element information, the template is determined based on the body information, and the element information is determined based on at least one of the sending object information and the objects to be recommended. The first sending module is used for feeding back the information to be displayed to a request main body corresponding to the request main body information.

According to an embodiment of the present disclosure, the apparatus further includes a storage device configured to store, before receiving the information request, a first mapping relationship between the request body information and information transmission associated information, where the information transmission associated information includes a template. Correspondingly, the information to be shown determining module comprises: the device comprises an object to be recommended acquisition submodule, a template determination submodule, an element information acquisition submodule and an assembly submodule. The object to be recommended acquisition submodule is used for acquiring one or more objects to be recommended aiming at the sending object information. The template determining submodule is used for determining a template according to the first mapping relation and the request main body information if the objects to be recommended are obtained and the number of the obtained objects to be recommended is larger than or equal to a number threshold. The element information acquisition submodule is used for acquiring the element information matched with the sending object information and the object to be recommended. The assembling submodule is used for assembling the template and the matched element information to obtain information to be displayed for the object to be recommended.

According to an embodiment of the present disclosure, the object to be recommended obtaining sub-module may include: the device comprises an object to be recommended acquiring unit and an object filtering unit. The object to be recommended acquiring unit is used for acquiring an object to be recommended. The object filtering unit is used for filtering the object to be recommended based on a preset rule to obtain a recommended object. Wherein the preset rule comprises at least one of: and whether the repeated acquisition associated information of the object to be recommended meets the repeated acquisition condition or not. Whether the object to be recommended belongs to a recommendation prohibition object or not, wherein the recommendation prohibition object comprises at least one of the following objects: the sending object forbids the recommending object, the terminal model forbids the recommending object, the request main body forbids the recommending object and the policy forbids the recommending object. Whether the inventory information of the object to be recommended meets the inventory condition or not.

According to the embodiment of the disclosure, the to-be-recommended object acquisition unit comprises a historical behavior commodity sequence information acquisition subunit and a first processing unit. The historical behavior commodity sequence information acquiring subunit is used for acquiring the historical behavior commodity sequence information corresponding to the sending object information. The first processing unit is used for processing the sending object information and the historical behavior commodity sequence information by using a first model to obtain information of an object to be recommended.

According to an embodiment of the present disclosure, the repeatedly acquiring the associated information includes a repeated purchase cycle. The device further comprises: and the second processing unit is used for inputting the sending object information and the information of the object to be recommended into the second model so as to obtain a repurchase period corresponding to the information of the object to be recommended.

According to an embodiment of the present disclosure, the apparatus further includes a second training module, where the second training module is configured to train the second model using training data such that an output of the second model approaches labeling information of the training data. The training data comprises sending object information, object information to be recommended, labeling information and at least one of the following: the commodity sequence information of the historical behaviors of the users, the commodity sequence information of the historical behaviors of the users in the same category and the commodity sequence information of the historical behaviors of all the users.

According to the embodiment of the present disclosure, the module for determining information to be displayed further includes: the hot spot object information acquisition sub-module and the post-processing sub-module. The hot object information obtaining submodule is used for obtaining information to be displayed of the hot object if the information to be displayed of the hot object is not obtained or the number of the objects to be recommended is less than a number threshold, wherein the information to be displayed of the hot object is matched with the information of the sending object. The post-processing submodule is used for taking the information to be displayed of the hot object as the information to be displayed corresponding to the information of the sending object, or complementing the information to be displayed of the object to be recommended by using the information to be displayed of the hot object, so that the quantity of the information to be displayed reaches the quantity threshold value.

According to an embodiment of the present disclosure, the apparatus further comprises: the system comprises a second storage module, a scheduling module, an auditing module and a scheduling determining module. Wherein the second storage module is configured to receive and store contract information prior to receiving an information request. And the scheduling module is used for generating scheduling information to be checked of the request main body based on the contract information. The auditing module is used for receiving an auditing result aiming at the scheduling information to be audited, and if the auditing result is passed, the scheduling information to be audited is used as scheduling information. The scheduling determining module is used for matching the request main body information and the request time with the scheduling information after receiving the information request, and responding to the information request if the matching is successful.

According to an embodiment of the present disclosure, the apparatus further comprises: the information sending device comprises an updating operation receiving module and a first updating module, wherein the updating operation receiving module is used for receiving the updating operation after storing a first mapping relation between request main body information and information sending correlation information. The first updating module is used for responding to the updating operation and updating the first mapping relation.

According to an embodiment of the present disclosure, the apparatus further comprises: the user operation receiving module and the second updating module. The user operation receiving module is used for receiving the user operation information after the request main body sends the information to be displayed to the main body sending the object information. And the second updating module is used for responding to the user operation information and updating the historical behavior commodity sequence information corresponding to the sent object information so as to at least update the object information to be recommended.

According to an embodiment of the present disclosure, the apparatus further comprises: and the load distribution module is used for responding to the information request and distributing the load after receiving the information request.

According to the embodiment of the disclosure, the load shunting module is specifically configured to perform load shunting in a common component mode. Each node of the server side is provided with an independent load balancer, and the load balancer supports single-node multi-instance deployment so as to distribute loads of multiple instances under the nodes conveniently.

According to the embodiment of the disclosure, the device adopts a multi-level cache, and the element information is stored in the multi-level cache. The multi-level cache comprises a server-side cache and a local cache, the server-side cache comprises at least two Redis clusters which are respectively used for storing at least part of the element information, and the server-side cache responds to a local request and transmits cached element information corresponding to the local request to the local cache.

Another aspect of the present disclosure provides an information processing apparatus including: the display device comprises an information request sending module, an information receiving module to be displayed and a second sending module. The information request sending module is used for sending an information request to the server side, wherein the information request comprises request main body information and sending object information. The information receiving module to be displayed is used for receiving information to be displayed aiming at the sending object from a server, wherein the information to be displayed aims at one or more objects to be recommended, the display information comprises a template and element information, the template is determined based on the main body information, and the element information is determined based on at least one of the sending object information and the objects to be recommended. The second sending module is configured to send the information to be displayed to the client of the sending object, so that the client of the sending object displays the information to be displayed.

According to the embodiment of the disclosure, the device further comprises a user operation information receiving module and a user operation information sending module. The user operation information receiving module is used for receiving the user operation information after the information to be displayed is sent to the client side of the sending object. The user operation information sending module is used for sending the user operation information to the server side so that the server side can update the historical behavior commodity sequence information corresponding to the sending object information and the information of the object to be recommended.

Another aspect of the present disclosure provides an electronic device comprising one or more processors and a storage, wherein the storage is configured to store executable instructions that, when executed by the processors, implement the method as described above.

Another aspect of the present disclosure provides a computer-readable storage medium storing computer-executable instructions for implementing the method as described above when executed.

Another aspect of the disclosure provides a computer program comprising computer executable instructions for implementing the method as described above when executed.

Drawings

The above and other objects, features and advantages of the present disclosure will become more apparent from the following description of embodiments of the present disclosure with reference to the accompanying drawings, in which:

fig. 1 schematically shows an application scenario of an information processing method, an information processing apparatus, an electronic device, and a medium according to an embodiment of the present disclosure;

FIG. 2 schematically shows a flow chart of an information processing method according to an embodiment of the present disclosure;

FIG. 3 schematically illustrates a schematic diagram of information to be presented according to an embodiment of the disclosure;

FIG. 4 schematically shows a schematic diagram of a repository, according to an embodiment of the present disclosure;

fig. 5 schematically illustrates a material center schematic according to an embodiment of the disclosure;

FIG. 6 schematically illustrates a logic diagram of an information processing method according to an embodiment of the present disclosure;

FIG. 7 schematically shows a flow chart of an information processing method according to another embodiment of the present disclosure;

FIG. 8 schematically illustrates a project management system according to an embodiment of the present disclosure;

FIG. 9 schematically illustrates a schedule management logic diagram according to an embodiment of the present disclosure;

FIG. 10 schematically shows a flow chart of an information processing method according to another embodiment of the present disclosure;

FIG. 11 schematically shows a schematic diagram of a playback engine according to an embodiment of the disclosure;

FIG. 12 schematically illustrates a logic diagram for load balancing according to an embodiment of the present disclosure;

FIG. 13 schematically illustrates an architecture diagram for load balancing according to an embodiment of the disclosure;

fig. 14 schematically shows a flow chart of an information processing method according to another embodiment of the present disclosure;

FIG. 15 schematically illustrates a logic diagram for data computation according to an embodiment of the present disclosure;

FIG. 16 schematically illustrates a logic diagram for a data query according to an embodiment of the present disclosure;

FIG. 17 schematically shows a flowchart for obtaining an object to be recommended according to an embodiment of the present disclosure;

FIG. 18 schematically shows a diagram of updating an object to be recommended according to an embodiment of the present disclosure;

fig. 19 schematically shows a block diagram of an information processing apparatus according to an embodiment of the present disclosure;

FIG. 20 schematically shows a block diagram of an information handling system according to an embodiment of the present disclosure;

fig. 21 schematically shows a block diagram of an information processing apparatus according to another embodiment of the present disclosure; and

FIG. 22 schematically shows a block diagram of an electronic device according to an embodiment of the disclosure.

Detailed Description

Hereinafter, embodiments of the present disclosure will be described with reference to the accompanying drawings. It should be understood that the description is illustrative only and is not intended to limit the scope of the present disclosure. In the following detailed description, for purposes of explanation, numerous specific details are set forth in order to provide a thorough understanding of the embodiments of the disclosure. It may be evident, however, that one or more embodiments may be practiced without these specific details. Moreover, in the following description, descriptions of well-known structures and techniques are omitted so as to not unnecessarily obscure the concepts of the present disclosure.

The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the disclosure. The terms "comprises," "comprising," and the like, as used herein, specify the presence of stated features, steps, operations, and/or components, but do not preclude the presence or addition of one or more other features, steps, operations, or components.

All terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art unless otherwise defined. It is noted that the terms used herein should be interpreted as having a meaning that is consistent with the context of this specification and should not be interpreted in an idealized or overly formal sense.

Where a convention analogous to "at least one of A, B and C, etc." is used, in general such a construction is intended in the sense one having skill in the art would understand the convention (e.g., "a system having at least one of A, B and C" would include but not be limited to systems that have a alone, B alone, C alone, a and B together, a and C together, B and C together, and/or A, B, C together, etc.). Where a convention analogous to "A, B or at least one of C, etc." is used, in general such a construction is intended in the sense one having skill in the art would understand the convention (e.g., "a system having at least one of A, B or C" would include but not be limited to systems that have a alone, B alone, C alone, a and B together, a and C together, B and C together, and/or A, B, C together, etc.). The terms "first", "second" and "first" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, features defined as "first", "second", may explicitly or implicitly include one or more of the described features.

The existing information flow advertisement putting flow and framework basically takes an advertisement position as a basic resource center, and the defects of the existing information flow advertisement putting flow and framework comprise that: the marketing plan is set up too long and has too many settings, such as time, title, description, etc.; secondly, enough manpower and time are needed to repeatedly prepare creatives and materials for different advertisement positions, a set of complete resources and a delivery plan management scheme is lacked, and the material reuse rate is low, so that the material resource management is extremely complicated. In addition, in the aspect of accurate delivery, the related art can use the same version of advertisement material for delivery, but is prone to cause aesthetic fatigue, cannot realize thousands of personalized deliveries even if the advertisement material is replaced regularly, and is high in material replacement cost. In addition, data collection is not real-time, and advertisements cannot be updated in real time following the current operation of the user.

Embodiments of the present disclosure provide an information processing method, an information processing apparatus, an electronic device, and a medium. The method includes an information determination process and an information feedback process. In the information determining process, in response to a received information request, information to be shown corresponding to sending object information is determined, wherein the information to be shown is specific to one or more objects to be recommended, the showing information comprises a template and element information, the template is determined based on the main body information, and the element information is determined based on at least one of the sending object information and the objects to be recommended. And after the information is determined, entering an information feedback process, and feeding back the information to be displayed to a request main body corresponding to the request main body information.

The information processing method provided by the embodiment of the disclosure has the advantages of uniform resource management, data precipitation and basic data service provision. The information processing method provided by the embodiment of the disclosure is based on scheduling configuration, improves marketing efficiency, reduces the operation complexity of personnel in a maintenance stage through pre-configuration, and is more automatic and intelligent. The information processing method provided by the embodiment of the disclosure is beneficial to realizing uniform media interfaces and improving the reusability of the interfaces due to uniform resource management. According to the information processing method provided by the embodiment of the disclosure, the delivery configuration is quickly formed by using the file schedule in the management stage, and the delivery schedule and the delivery related content are efficiently established. According to the information processing method provided by the embodiment of the disclosure, the material resource is templated, the accuracy of material specification is improved, and the improvement of the material reuse rate is facilitated. The information processing method provided by the embodiment of the disclosure adopts a deep learning technology and combines with stream computing to update the historical behavior information of the user in real time, dynamically adjusts the recommendation strategy according to the real-time operation of the user, and simultaneously avoids repeatedly recommending commodities in a non-repurchase period. The information processing method provided by the embodiment of the disclosure adopts the corresponding media template to fill, play and display the content, and ensures that each media of the same content is multiplexed. The information processing method provided by the embodiment of the disclosure ensures that users can contact the latest and different contents at the same time, different media, different channels of the same media and different advertisement spots of the same channel through the play frequency control.

Fig. 1 schematically shows an application scenario of an information processing method, an information processing apparatus, an electronic device, and a medium according to an embodiment of the present disclosure. It should be noted that fig. 1 is only an example of a system architecture 100 to which the embodiments of the present disclosure may be applied to help those skilled in the art understand the technical content of the present disclosure, and does not mean that the embodiments of the present disclosure may not be applied to other devices, systems, environments or scenarios.

As shown in fig. 1, the system architecture 100 according to this embodiment may include

terminal devices

101, 102, 103, a network 104, a server 105 and a media side 106. The network 104 is used to provide a medium for communication links between the

terminal devices

101, 102, 103, the server 105 and the media side 106. Network 104 may include various connection types, such as wired, wireless communication links, or fiber optic cables, to name a few.

The user may use the

terminal devices

101, 102, 103 to interact with the server 105 via the network 104 to receive or send messages or the like. The

terminal devices

101, 102, 103 may have installed thereon various communication client applications, such as shopping-like applications, web browser applications, search-like applications, instant messaging tools, mailbox clients, social platform software, etc. (by way of example only).

The

terminal devices

101, 102, 103 may be various electronic devices having a display screen and supporting advertisement playing, including but not limited to smart phones, tablet computers, laptop portable computers, desktop computers, and the like.

The server 105 may be a server providing various services, such as a background management server (for example only) providing support for websites browsed by users using the

terminal devices

101, 102, 103. The backend management server may analyze and perform other processing on the received data such as the user request, and feed back a processing result (for example, a web page, information, or data obtained or generated according to the user request, or a maintenance result of maintaining a material library and the like according to the user operation) to the terminal device. The server 105 may also generate and transmit advertisements to the media side 106 in response to an advertisement request from the media side 106.

The media terminal 106 can deliver the advertisement to the

terminal devices

101, 102, 103, analyze the effect of the delivered advertisement, and feed back the operation of the user for the advertisement to the server 105 for data update and the like.

It should be noted that the information processing method provided by the embodiment of the present disclosure may be executed by the server 105. The information processing method provided by the embodiment of the present disclosure may also be executed by a server or a server cluster that is different from the server 105 and is capable of communicating with the

terminal devices

101, 102, 103 and/or the server 105.

It should be understood that the number of terminal devices, networks, and servers are merely illustrative. There may be any number of terminal devices, networks, and servers, as desired for implementation.

Fig. 2 schematically shows a flow chart of an information processing method according to an embodiment of the present disclosure.

As shown in fig. 2, the method includes operations S201 to S205.

In operation S201, an information request including request body information and transmission object information is received.

For example, if the media side needs to send an advertisement to a designated terminal device (the terminal device has a mapping relationship with the sending object information), the identification information of the media itself, the sending object information (such as a user identification) and the like can be sent to the server side, so that the server side can determine a commodity to be recommended according to the user preference and the like corresponding to the user identification, can determine a template of the advertisement according to a contract, an advertisement space and the like, determine a material according to the commodity to be recommended, the model of the terminal device and the like, and facilitate automatic assembly of the advertisement required by the media side.

Then, in operation S203, in response to the information request, information to be presented corresponding to the sending object information is determined, where the information to be presented is for one or more objects to be recommended, the presentation information includes a template determined based on the subject information and element information determined based on at least one of the sending object information and the objects to be recommended.

The object to be recommended may be determined based on the following manner.

For example, the cold start problem can be effectively solved by defining the popularity based on the recommendation of the popularity and according to the browsing amount, the sales amount, the social hot spot, and the like. As another example, based on collaborative filtering recommendations, relying only on user behavior, there is no need for deep understanding of commodity features or metadata. The coverage rate is high, the problem of long tail can be effectively solved, and the user is surprised. As another example, content/rule/knowledge-based recommendations rely on a complete, well-categorized content/rule/knowledge hierarchy and structured data features. The timeliness is high, the recommendation reason is reasonably explained, and the user trust degree is high. For another example, based on the combination of the above recommended algorithms, advantages are complementary, cold start, data sparseness and data unstructured problems can be avoided, and the application range is wide. In addition, the object to be recommended may be adjusted based on a preset rule or the like. For convenience of understanding, the following description will take the object to be recommended as a recommended product as an example.

In order to improve the convenience of the maintenance and use stage, the advertisement position which can be used by the media can be determined in advance according to the contract signed with the media, the applicable template can be determined according to the advertisement position, and then the adaptive element information (such as materials and the like) can be determined according to the object to be recommended, the template and the like. In addition, the materials can be further screened according to parameters of the terminal for displaying the information to be displayed and the like, so that the optimal display effect is achieved. Thus, the template and element information can be assembled into the advertisement needed by the media.

In operation S205, the information to be shown is fed back to the request subject corresponding to the request subject information.

In one embodiment, the method may further comprise the operations of: before receiving an information request, storing a first mapping relation between request body information and information transmission associated information, wherein the information transmission associated information comprises a template.

Accordingly, the determining of the information to be presented corresponding to the transmission object information may include the following operations.

First, one or more objects to be recommended for the transmission object information are acquired. Such as predicting goods that meet the user's desire to purchase.

Then, if the objects to be recommended are acquired and the number of the acquired objects to be recommended is greater than or equal to a number threshold, determining a template according to the first mapping relation and the request subject information.

And then, acquiring element information matched with the sending object information and the object to be recommended. If the object to be recommended is a single lens reflex, the adaptive pictures and/or texts (such as the name, parameters, brief introduction and the like of the single lens reflex, which can be provided by the media party, and can also be found in the history database) can be inquired from the material library.

And then assembling the template and the matched element information to obtain the information to be displayed for the object to be recommended.

In another embodiment, the method may further include the following operations.

And if the objects to be recommended are not acquired or the number of the objects to be recommended is less than the number threshold, acquiring the information to be displayed of the hot object, wherein the information to be displayed of the hot object is matched with the information of the sending object.

Then, the information to be displayed of the hot object is used as the information to be displayed corresponding to the information of the sending object, or the information to be displayed of the hot object is used for complementing the information to be displayed of the object to be recommended, so that the quantity of the information to be displayed reaches the quantity threshold value.

Fig. 3 schematically shows a schematic diagram of information to be presented according to an embodiment of the present disclosure.

As shown in fig. 3, the top left of fig. 3 is a template, and may include a picture template and a text template (for the advertisement field, it may also be referred to as a pattern template). The picture template and the text template can be independent templates or combined templates. The template can specify the position, range, plate type, transparency and other attributes of the picture, the background, the position, range, font size, color and the like of characters, wherein the material can be searched from a resource library (such as image information and character information) or sent by an information requester (such as character information). The lower left diagram of fig. 3 is element information, such as material, which may be specifically picture material, text material, and the like. After the adapted template and material are determined, the template and material may be assembled into an advertisement as shown in the right diagram of fig. 3. The advertisement shown in the right diagram of fig. 3 may then be sent to the media side as information to be presented, whether to send the advertisement to a specified terminal device may be decided by the media side, and so on.

FIG. 4 schematically shows a schematic diagram of a repository, according to an embodiment of the present disclosure.

As shown in FIG. 4, the repository may support resource management and project management. For example, base data may be maintained to provide data services for managing scheduling configurations.

Specifically, the correlation with respect to resource management may be as follows.

Media type management: media type information. There are add, modify, query functions.

Media management: media information, media type information. There are add, modify, query functions.

And (3) channel management: media type information, media information, channel information. There are add, modify, query functions.

Material specification management: media information, material size, material copy. There are add, modify, query functions.

And (3) purchasing type management: and (4) purchasing type information. There are add, modify, query functions.

And (3) advertisement position management: media type information, media information, channel information, material specification information, purchase type information, and the like. There are add, modify, query functions.

Managing a picture template: the material picture template size, picture type and size are set. There are add, modify, query functions.

And (3) managing the document template: the material file template has fine characters. There are add, modify, query functions.

Managing an active pool: the method comprises the steps of setting information such as a movable landing page, a movable label, a movable material, a movable case and the like. There are add, modify, query functions.

The correlation with respect to project management may include the following.

Managing by an agent company: and maintaining information such as agent companies, agent accounts and the like. There are add, modify, query functions.

Channel management: tertiary channel, quaternary channel, etc. There are add, modify, query functions.

And (3) putting brand management: and releasing brand type information. There are add, modify, query functions.

And (4) using the project: and maintaining the project expense owner information.

And (3) project responsible person management: project principal information and member information. There are add, modify, query functions.

And (4) project management: project information, agent companies to which the projects belong, third-level channel information, fourth-level channel information, release brand information, project user information, project leader information, project start and stop time, project cost and the like. There are add, modify, query functions.

The resource library supports the unified management of basic resources, and a user inputs related information such as an agent, a channel, media, an advertisement position and the like, so that data service support is provided for a management system and a playing system. Therefore, in the maintenance and use stage, the information such as the advertisement position, the template and the like can be automatically determined only by inputting the identification information of media and the like, and the convenience of use is improved. In addition, the resource reuse is facilitated.

Fig. 5 schematically shows a material center schematic according to an embodiment of the present disclosure. As shown in fig. 5, the feedback of the information to be displayed to the request subject can be realized by the material center and the playing engine. Specifically, the material center bears the production and management of the advertisement material, processes the basic data to generate the advertisement material, and writes the advertisement material into a cache for a playing engine to use after formatting according to a specified template structure. The playing engine interacts with the media in a mode of exposing a hypertext transfer protocol (HTTP) interface to realize advertisement putting: and after receiving the media advertisement request, the playing engine reads the material cache and calls an interface to obtain a recommended commodity list, and returns the media advertisement after performing related calculation and assembly for media exposure display.

For example, the material center mainly includes 5 portions of project creation planning, creation unit, binding delivery scheduling, creation creative, binding template and activity, and material caching, and depends on the project, picture template, document template, activity pool of the resource library shown in fig. 4 and the delivery scheduling of the management system. The material center can output the material and write the material into the cache for the playing engine to use by executing the following 6 operations.

Fig. 6 schematically shows a logic diagram of an information processing method according to an embodiment of the present disclosure. As shown in fig. 6, the above 6 operations are as follows.

And (3) according to a project new project: and calling a project list from the resource system, selecting a certain project to create a release plan, wherein the release plan time span is contained in the project time span, and a plurality of plans can bind the same project.

Newly building a unit: a new unit is planned, the proportion of unit deselection amount (namely, a certain proportion of advertisement exposure opportunities are abandoned) is set, and a plurality of units can be built under the same plan.

Binding release scheduling: on the unit, according to the project and time span of the plan, the schedule list is pulled from the schedule system, and the proper schedule is selected for binding, thereby setting the advertisement position and date which can be used for putting by the current unit.

Newly creating an original: a specific ad slot scheduling newly-built creative idea is specified under a unit, wherein the creative idea comprises an accurate creative idea and an activity creative idea (namely a bottoming creative idea), the accurate creative idea corresponds to the release of recommended commodities, and the activity creative idea corresponds to the condition that the recommended commodities are not obtained.

Binding templates and activities: and the accurate creativity is that the picture template and the document template are pulled from the resource system, and the templates meeting the specifications of the advertising positions to which the picture template and the document template belong are selected for binding so as to generate the accurate material template. The activity creativity is bound with the activity matched with the specification and the schedule of the advertisement space to which the activity creativity belongs (the activity can be used as long as the pictures and the documentations of the noble sides of the advertisement space exist), so that the activity material can be generated.

Caching materials: based on the generated creatives, screening and assembling of necessary attributes are carried out, and advertisement materials are generated and stored into a Redis (high-performance key-value database, wherein the Redis cluster is composed of a plurality of nodes (fragments)) cluster for real-time reading of a playing system. And completing a whole set of flow at the material center, and waiting for the playing engine to read the material for playing.

Fig. 7 schematically shows a flowchart of an information processing method according to another embodiment of the present disclosure. The following describes the project management with reference to fig. 7.

As shown in fig. 7, the method may further include the following operations.

In operation S701, contract information is received and stored before an information request is received. The contract information may be requirements, rules, etc. agreed in contracting with the media.

In operation S703, to-be-audited scheduling information of the request subject is generated based on the contract information. The scheduling information to be audited may be a scheduling generated based on contract information, which requires auditing to reduce an error rate, and the like.

Operation S705 is performed to receive an audit result for the scheduling information to be audited, and if the audit result is that the scheduling information to be audited passes, the scheduling information to be audited is used as the scheduling information.

Operation S707, after receiving the information request, matches the request body information and the request time with the scheduling information, and if the matching is successful, responds to the information request. Such as determining whether the request body information and the request time included in the information request are matched with the body information and the output time corresponding to the schedule.

FIG. 8 schematically shows a project management system according to an embodiment of the disclosure.

As shown in FIG. 8, project management may include resource setup, schedule upload, schedule review, schedule modification, and monitoring combing.

Wherein, the resource setting: project management needs to rely on system settings. Resource advertisement position information and project information required by scheduling are set in the system, and the resource advertisement position information and the project information comprise information such as media types, media, channels, advertisement positions, purchase types, material specifications, picture templates, file templates, activity pools, agents, project users, responsible persons, delivered brands, channels, projects and the like.

Scheduling uploading: uploading a scheduling template and uploading a monitoring carding table. For example, a dedicated Excel template can be set, which can be downloaded. Each schedule is accurate to the corresponding ad slot and the schedule time that the ad slot needs to be delivered. The data of the same item and the same advertisement space are unique. The range of the scheduling time takes the item start-stop time as the start-stop range. The date of release required can be selected within this range. From above the data structure can be understood as a "row table" made with the scheduling time. Specifically, the schedule table shown in table 4 can be referred to.

TABLE 1

Table 1 shows that on the same project and at the same slot, there is a schedule of impressions on all four days 2018.10.10 through 2018.10.13.

And (4) scheduling audit: and the manager reviews the scheduling information. Including the verification of scheduling resource advertisement space information and project related information. After the verification is passed, the manager can review the schedule record. And after the audit is passed, "pass", and if the audit is not passed, "reject", filling the reason for not passing the audit.

Scheduling modification: the scheduling may be modified or deleted. Ad slots and item related information in the listing are not allowed to be modified. And after the modification is successful, the state is changed into a state to be audited. The manager can review the information.

Monitoring and carding: and after the scheduling audit is passed, generating corresponding monitoring carding table information aiming at the scheduling of each day. It is understood that the scheduling time is broken down into a "list". For example, the schedule of table 1, the generation of the monitoring comb table becomes 4 pieces of data. 2018.10.10, 2018.10.11, 2018.10.12, 2018.10.13, respectively. The scheduling information comprises monitoring links spliced by all information in the scheduling according to the scheduling information and finally released monitoring links. The data in the monitoring carding table is real scheduling data used for putting and reporting data.

In another embodiment, the method may further include the following operations.

After storing the first mapping relationship between the request subject information and the information transmission associated information, an update operation is received.

Then, in response to the update operation, the first mapping relationship is updated.

FIG. 9 schematically illustrates a schedule management logic diagram according to an embodiment of the present disclosure.

The scheduling management logic shown in fig. 9 determines the basic configuration of the entire ad placement, i.e., which agent is operating, on which channel and specific ad slot of the media, when the placement is performed (when the media calls the play interface).

Specifically, 1, resource advertisement space information is set, and corresponding resource information is required to be established in media type management, media management, channel management, purchase type management, material specification management, picture template management, file template management and activity management in a database related to resource management. And then newly building an ad slot in the ad slot management. The ad spot management data is entered into the Mysql database. And newly building agent companies, channel information, released brand information, project users and project responsible persons in a database related to project management. And newly building a project in project management, and inputting project information. And the related information of the project is recorded into a Mysql database. These resource ad slot information may be updated.

2. The user logs in the system, downloads the scheduling template table and inputs scheduling basic information into the scheduling table, wherein the scheduling basic information comprises resource advertisement position information and project information. When the resource advertisement space information and the project information are acquired, the acquired data are temporarily cached in Redis. And the basic data acquisition efficiency is improved.

3. And filling relevant expenses, key performance indicator KPI pre-estimated values (such as click, exposure, cost of thousands of people CPM, advertisement CPC paid per click and click through rate CTR) and scheduling time into a scheduling table. The scheduling time is stored in Redis and a non-relational database (such as an elastic search, ES for short) to generate complete scheduling table information.

4. Uploading a scheduling table, and verifying an information system in scheduling.

5. And if the scheduling system fails to check, prompting that the scheduling information is wrong, and filling the wrong data into the auditing remarks in the scheduling table. The user can obtain the failing prompt information by downloading the schedule through the page.

6. And adjusting scheduling data according to the prompt and re-uploading.

7. And the scheduling system passes the verification and the scheduling information is put in storage. The scheduling state is changed into 'to be checked', and scheduling information can be inquired on a checking page.

8. And if the scheduling audit is not passed, recording the rejection reason, and changing the scheduling state into 'rejection'. And warehousing the state change data.

9. The scheduling audit is passed, and the scheduling state is changed into 'pass'. And (4) entering state change data into a database, and synchronously generating monitoring and combing information, wherein the monitoring and combing information comprises resource advertisement space information, project related information, scheduling cost and KPI (Key performance indicator) pre-evaluation value. And generating a scheduling monitoring link and a final release link through the information. And generating a monitoring carding table, and warehousing monitoring carding data.

10. And after the monitoring combing data is generated, the monitoring combing data is used for monitoring the advertisement sending state and the like.

11. And the scheduling information is modified on line, and only the cost, KPI estimation and scheduling time parameters can be modified.

12. And deleting the scheduling information on line.

13. And after online modification or deletion, entering modified data verification, passing the verification, changing the scheduling state into 'to be audited', and warehousing the data. And if the verification fails, the page prompts the reason.

Fig. 10 schematically shows a flowchart of an information processing method according to another embodiment of the present disclosure.

As shown in fig. 10, the method further includes operation S1001.

In operation S1001, load shedding is performed in response to an information request after the information request is received.

Specifically, the load splitting includes splitting the load in a common component mode. Each node of the server side is provided with an independent load balancer, and the load balancer supports single-node multi-instance deployment so as to distribute loads of multiple instances under the node.

In another embodiment, the element information employs a multi-level cache. The multi-level cache comprises a server-side cache and a local cache, the server-side cache comprises at least two Redis clusters which are respectively used for storing at least part of the element information, and the server-side cache responds to a local request and transmits cached element information corresponding to the local request to the local cache.

Fig. 11 schematically shows a schematic diagram of a playback engine according to an embodiment of the present disclosure.

As shown in fig. 11, the playback engine includes a material cache, a local cache of the material, a pre-checker, an advertisement calculator, an advertisement assembler, etc., and relies on a common component load balancer, a commodity recommendation interface of the commodity recommendation system.

Fig. 12 schematically illustrates a logic diagram for load balancing according to an embodiment of the present disclosure.

As shown in fig. 12, the playback engine returns the information to be presented to the media for exposure. The load distribution is completed by a common component Linux Virtual Server (LVS), and the commodity list is obtained by calling a commodity recommendation system interface, as shown in the following.

Media request: the operator can put the list under the media side in advance, and put the corresponding ad slot ID synchronously, after reaching putting time, the media uses the ad slot ID and relevant parameters, calls the external playing interface of the playing engine, and requests the advertisement in real time.

Load shunting: and the LVS (a load balancer of a 4-layer (transmission layer tcp/udp) is used, each function module of the LVS is built in a version behind Linux2.4) to perform load distribution, and the media request is distributed to each back-end service to perform service logic processing.

Pre-checking: after receiving the advertisement request, the playing engine firstly enters a preposed checker to check the legality of the request parameter, if the advertisement position ID is checked to be empty, the next flow is entered if the check is passed, otherwise, the parameter error is directly returned.

Reading and caching materials: the ad material is read into a material cache (i.e., a material cache in the material center) in conjunction with the media request ad slot ID, the current request time, and other relevant request parameters (e.g., size).

Calling a recommended commodity: based on the media request ad slot ID, current request time, request device number (which can be mapped to a specific user), the RPC interface of the product recommendation system is invoked (remote procedure call, which is a protocol that requests services from a remote computer program over a network without knowledge of underlying network technology), a list of recommended products is obtained and entered into the ad calculator.

And (3) advertisement calculation: and performing frequency control calculation based on the obtained commodity list, screening out a certain commodity needing to be recommended currently, and calculating and generating a monitoring link, a picture link and a file content for putting in combination with the read material cache. If the product list is empty, the advertisement calculator reads the active material from the material cache for the priming (if the backlog is generated, the advertisement calculator is circulated to the advertisement assembler to assemble the backlog message body).

Advertisement assembly and return: and assembling advertisement related data (such as templates and materials) according to a specific data protocol (Protobuf) and a data structure to generate the media advertisement. The assembled and generated media advertisement (including the display picture, the promotion file information and the monitoring link) is returned to the media side through the external playing interface, and the whole playing process is completed.

Fig. 13 schematically illustrates an architecture diagram for load balancing according to an embodiment of the present disclosure.

As shown in fig. 13, in order to support a large number of off-site advertisement requests, a series of optimizations are performed on the broadcast system architecture to ensure the stability and response timeliness of the system. .

Specifically, a public component (Linux Virtual Server, LVS for short) DR mode can be used for load distribution, which has strong load resistance, high performance and low consumption on memory and CPU resources. And the media request is distributed to each service end node through the LVS for service logic processing. The LVS is also provided with an intelligent monitoring mechanism, unhealthy machines can be automatically removed, and the balance calling of all instances in the cluster is transparently completed. And various load strategies such as polling, weighted polling, polling + the same machine room and polling + the same network segment are supported, configuration can be carried out according to specific scenes, optimal flow distribution is achieved, and the overall performance of the cluster is improved.

In addition, each service end node can have its own independent nginx (high-performance HTTP and reverse big-key server, also can be used as load balancer), and can make domain name resolution or correspondent error redirection treatment, etc. and can support single-node multi-instance deployment, at this moment, nginx can implement load distribution of every instance under the current node.

In the playing system, a multi-level cache mechanism is adopted for the material, and the multi-level cache mechanism comprises a server cache (stored in a Redis cluster) and a local cache, which are also called a first-level cache and a second-level cache.

The first-level cache can adopt double Redis cluster deployment, the current used cluster is determined by module taking according to key, the flow is halved to two clusters, and the probability of cluster single-chip overheating or single-cluster link number excess is reduced.

The second-level Cache (e.g. Loading Cache, a local Cache mechanism of *** open source, supporting dynamic Cache refresh), establishes a dynamic Cache between the business logic and the Redis, stores the material content obtained from the first-level Cache into the local, and directly reads the material content from the local next time, which is essentially a local Cache of the business program. The second-level cache also supports cache asynchronous penetration refreshing, cache number control, low-frequency cache removal in excess priority and other mechanisms. As a buffer area before Redis, the secondary cache can effectively improve the service processing speed, protect the Redis cluster and effectively reduce the load of the Redis cluster.

Fig. 14 schematically shows a flowchart of an information processing method according to another embodiment of the present disclosure.

As shown in fig. 14, the method further includes operations S1401 to S1403.

In operation S1401, after the request body transmits information to be presented to the body transmitting the object information, user operation information is received.

Specifically, the user operation information may be an operation performed by the user at the terminal device for the advertisement. Such as clicking to view the advertisement, purchasing goods displayed in the advertisement, consulting, etc.

In operation S1403, in response to the user operation information, historical behavior commodity sequence information corresponding to the sending object information is updated, so that at least the to-be-recommended object information is updated.

Since there may be an action of purchasing a product or the like after the user views the advertisement, the action should be linked with the object to be recommended. For example, if a user purchases a certain item, the user is less likely to purchase the item again in a short period of time, and the item is recommended to the user again in a short period of time, the user may waste advertisement resources, and the user may be classified as useless information, which may degrade the user experience.

In order to enable the real-time operation of the user to be linked with the object to be recommended, the linkage can be realized in a streaming computing mode, and a high-concurrency scene can be supported.

Specifically, the stream-oriented computation mainly collects the information in the station and the statistical data provided by the real-time computation in real time and performs cleaning: orders, browsing, inventory, etc., and incorporate historical data for calculations. And the calculation adopts a deep learning algorithm and reinforcement learning to carry out real-time recommendation for the user. And returning the latest calculation commodity and the picture thereof in real time when the system requests. The stream computing mainly comprises the following steps: a data calculation process and a data query process.

FIG. 15 schematically shows a logic diagram for data computation according to an embodiment of the disclosure.

As shown in fig. 15, the data calculation may include the following operations.

Receiving real-time data: and the exposure browsing log, the real-time order log, the real-time inventory message and the like are butted through a Kafka (distributed publishing and subscribing message system) and introduced into a Storm (distributed real-time computing system) cluster of a streaming system.

Data cleaning: filtering a no user name (Pin) log, and keeping the Pin log for association at last.

And (3) data filtering: and filtering the commodities prohibited to be thrown.

Vector matching: the cleaned data is mapped to the corresponding product Uniform number (SKU) degree of the user and the specific commodity.

Counting and counting: and counting the data according to the dimensions of the SKU and the pin as a parameter weight.

Model vector: and processing the sorted data into vectors of Pin + SKU + SKU browsing times + SKU information according to a model, and combining the vectors together with the Pin vector of the Pin information.

DDPG model calculation: and performing model calculation according to the vector data and the historical data in the last step to obtain the latest recommended commodity of the current pin, scoring and converting the latest recommended commodity into a result Redis.

FIG. 16 schematically illustrates a logic diagram for data querying in accordance with an embodiment of the present disclosure.

As shown in fig. 16, the data query may include the following operations.

Equipment number inquiry: triggered by the playing system, using the equipment information to call a streaming algorithm; and querying a result set in Redis according to the mapping of the equipment to the user Pin dimension by using the streaming algorithm.

And (4) filtering results: the result set is filtered in real time, including the latest SKU inventory, the latest contra item SKU level, and the order, i.e., whether or not a purchase has been made.

Recommending compensation: and performing SPU-level deduplication on the filtered sum result, and supplementing the hottest commodities for filling if the number is not enough than a threshold value, so as to ensure the diversity viewed by the final user.

And (4) sequencing results: and sorting the final commodity results according to the scoring results of the learning algorithm, outputting the final commodity results to a playing system, and carrying out frequency control by the playing system.

In another embodiment, in the method for acquiring an object to be recommended as described above, including recommendation based on popularity (popularity), recommendation based on collaborative filtering, recommendation based on content/rule/knowledge, and a combination of these recommendation algorithms, since the recommendation result is output to the user for presentation, and no dynamic adjustment is performed on the recommendation policy according to the user feedback (such as whether to click or not, whether to purchase, etc.), the recommendation effect is not ideal. In addition, due to the fact that the specific repurchase cycle factor of the commodity is not considered (generally, the repurchase cycle of the consumable part is larger than that of the fast-consumed part), repeated recommendation of the commodity in a non-repurchase cycle is caused, and resource waste of the display position and user experience problems (such as contradictory conflict, distrust and the like) are caused.

Fig. 17 schematically shows a flowchart of acquiring an object to be recommended according to an embodiment of the present disclosure.

As shown in fig. 17, the acquiring one or more objects to be recommended for the transmission object information may include operations S1701 to S1703.

In operation S1701, an object to be recommended is acquired.

In one embodiment, the obtaining of the object to be recommended may include: first, historical behavior commodity sequence information corresponding to transmission object information is acquired. For example, user information, commodity information, and commodity sequence information of user historical behaviors are acquired.

The acquired user information, commodity information and commodity sequence information of the user historical behaviors are acquired and processed by stream computing. Specifically, user information, commodity information, user behavior information and user order information of each user are collected, and then the collected information is sequentially subjected to extraction, cleaning, characterization and labeling to obtain the processed user information, commodity information and user historical behavior commodity sequence information.

The user information includes, but is not limited to, at least one of: user rating, gender, user age, user marital status, academic history, occupation, presence or absence of a car, the method comprises the following steps of firstly purchasing the garment in the same year, last-year shopping, first single time to present, present time, last login present time, first order present days in the same year, last month unit price, last two month unit prices, last three month unit prices, mother and baby medal level, individual care and make-up medal level, wine medal, user liveness model, purchasing power segmentation, life cycle, RFM full-quality category score, POP garment RFM grouping, POP garment RFM standardized score, POP garment RFM super garment RFM grouping, POP garment super RFM standardized score, large garment quality RFM grouping, large garment quality RFM standardized score, user value grouping and user value standard score.

After the user behavior information and the user order information are processed, the commodity sequence information of the user historical behaviors is obtained, and the commodity sequence information comprises but is not limited to at least one of the following: a user history browsing merchandise ID sequence, a user history purchasing merchandise ID sequence, a user history exposure merchandise ID sequence, etc.

The commodity information is processed to obtain commodity price, size, brand, sales volume, class and the like. The commodity here refers to a commodity obtained from commodity sequence information of user historical behaviors.

Specifically, when the current user browses the advertisement and performs online access, the related information of the current user is searched, and a to-be-recommended commodity list of the current user is obtained according to the first model. For example, when the current user accesses, the user information, the commodity information and the commodity sequence information of the historical behavior of the user are obtained according to the current user identifier. And then inputting the user information, the commodity information and the commodity sequence information of the historical behaviors of the user into a first model to obtain a to-be-recommended commodity list of the current user.

In a particular embodiment, the first model includes a predictor submodel and a parameter update submodel.

The input of the predictor model comprises sending object information and historical behavior commodity sequence information, the historical behavior commodity sequence information comprises target marking information, and the output of the predictor model comprises a recommended object and a corresponding accuracy prediction result.

The input of the parameter updating submodel comprises an accuracy prediction result and target labeling information corresponding to the sending object information, and the output of the parameter updating submodel comprises an optimization model parameter, wherein the optimization model parameter is obtained based on a first loss function, and the first loss function is constructed based on the accuracy prediction result and the target labeling information corresponding to the sending object information. The predictor sub-model may be an eval neural network. The parameter update submodel may be a target nerve.

In one embodiment, the first model may be trained using reinforcement learning.

Specifically, firstly, two identical neural networks, namely an eval neural network and a target neural network, are constructed, wherein the eval neural network is used for obtaining a recommended commodity list and a predicted value of the quality degree of a recommended result, and the target neural network is used for updating parameters of the eval neural network.

For example, eval neural network is used to obtain the recommended goods list (action) and the predicted value Q of the quality of the recommendation result.

Initializing eval neural network parameters θ^QAnd theta^μ(ii) a Initializing target neural network parameters θ^Q′＝θ^Q，θ^μ′＝θ^μ。

Using the user information, commodity information and commodity sequence information of the historical behaviors of each user as a piece of training data s_i. Will(s)_i,a_i,r_i,s_i+1) As the ith sample in the N samples of the reinforcement learning training set. Wherein, a_iList of recommended items for ith sample, r_iUser feedback for the ith sample, s_i+1Is the ith sampleUser information, commodity information and commodity sequence information of user historical behaviors in the next state. Wherein(s)_i,a_i,r_i,s_i+1) As a sample, by_iBehavioral influence is given by s_i+1. By analogy, the i +1 th sample is(s)_i+1,a_i+1,r_i+1,s_i+2)。

And then, constructing a first loss function according to the target value of the recommendation result quality degree of the ith sample and the prediction value of the recommendation result quality degree. In the eval neural network, the first loss function is optimized by taking a small value, and the parameter theta of the eval neural network is updated^Q。

Wherein the optimization model parameter corresponds to a direction in which a gradient of the first loss function decreases fastest, and corresponds to the first loss function taking a small value. For example, the gradient of the first loss function is calculated by a gradient descent method, and the direction in which the gradient descends most quickly is selected so that the first loss function is minimized. In the eval neural network, the expectation function fed back by the user is optimized by taking a small value, and the parameter theta of the eval neural network is updated^μ. Thus, the value of θ can be determined^QAnd theta^μUpdating the parameter of the target neural network to be theta^Q′And theta^μAnd obtaining the trained first model.

And then, processing the sending object information and the historical behavior commodity sequence information by using a first model to obtain the information of the object to be recommended.

In operation S1703, the object to be recommended is filtered based on a preset rule, so as to obtain a recommended object.

Wherein the preset rule comprises at least one of: and whether the repeated acquisition associated information of the object to be recommended meets the repeated acquisition condition or not. Whether the object to be recommended belongs to a recommendation prohibition object or not, wherein the recommendation prohibition object comprises at least one of the following objects: the sending object forbids the recommending object, the terminal model forbids the recommending object, the request main body forbids the recommending object and the policy forbids the recommending object. Whether the inventory information of the object to be recommended meets the inventory condition or not.

In one embodiment, the repeatedly acquiring the associated information includes a repurchase period.

Accordingly, the method further comprises: and inputting the sending object information and the information of the object to be recommended into a second model so as to obtain a repurchase period corresponding to the information of the object to be recommended.

Specifically, the historical purchase recording time of the current user can be compared with the re-purchasing period of each commodity in the commodity list, and the recommended commodity list of the current user is obtained after filtering the non-re-purchased commodities in the commodity list. Therefore, the cache of the recommended commodity result is completed and is continuously updated, and the latest result is obtained by adopting a query stream type calculation query flow during advertisement playing.

Fig. 18 schematically shows a schematic diagram of updating an object to be recommended according to an embodiment of the present disclosure.

As shown in fig. 18, the real-time calculation mainly receives real-time data and performs real-time processing, and the Storm scheme provides second-level data processing, and the highest throughput peak value 80w/s in a high concurrency period (such as a jingdong 618 promotion period, a jingdong double eleven promotion period, and the like) is used for providing timeliness verification for ensuring advertisement delivery for users, and tracking delivery process changes and data trends for delivery decision support.

Specifically, real-time messages are received, including but not limited to: real-time request, exposure, click, browse, order.

The real-time message is parsed to determine the advertisement slots to which the message pertains.

And the advertisement space information is calculated in real time and stored in the Redis cluster for inquiry.

The cached data is stored in Hbase (high-performance big data key-value storage system, which can store data persistently) for persistence.

Data details in Hbase can be displayed in a summary mode. And updating the resource library through streaming calculation based on the data in the Hbase, and further updating algorithm model data and the like based on the data in the resource library so as to update the user recommended materials (including but not limited to recommended commodities, materials and the like).

In one embodiment, the second model may be trained in the following manner.

Training the second model with training data such that the output of the second model approaches the labeled information of the training data. The training data comprises sending object information, object information to be recommended, labeling information and at least one of the following: the commodity sequence information of the historical behaviors of the users, the commodity sequence information of the historical behaviors of the users in the same category and the commodity sequence information of the historical behaviors of all the users.

In one implementation, the second model may be a deep learning neural network. The training process can be as follows.

And taking the user information of the user and the commodity information bought by the user as the ith sample in M samples of the deep learning training set, wherein i belongs to M, and M is a natural number.

And inputting the user information of the ith sample and the commodity information bought by the user into the second model to obtain a commodity re-purchasing period training value.

And constructing a second loss function according to the commodity repurchase cycle training value and the repurchase cycle real value of the ith sample, optimizing the minimum value of the second loss function, and updating the network weight parameter to obtain a trained second model.

Wherein the real value of the re-purchasing period of each sample is determined by the real purchasing record time. When the training value of the sample is closer to the true value, the second loss function takes the minimum value, the obtained network weight parameter is optimal, the model training is finished, and the trained second model can be used for determining a commodity repurchase cycle for the current visiting user in the subsequent step 3, namely, the repurchase cycle of the current user for each commodity in the commodity list is obtained.

And calculating the gradient of the second loss function by adopting a gradient descending method, and selecting the direction with the fastest gradient descending so as to enable the second loss function to be the minimum. The second model is trained and corresponds to a set of optimal network weight parameters.

The first model and the second model may be models with learning capabilities, including but not limited to a variety of neural networks.

The algorithm model provided by the embodiment of the disclosure is mainly characterized in that: the recommendation strategy can be dynamically adjusted according to the user feedback while avoiding repeated recommendation of commodities in a non-repurchase period. The core idea is as follows: first, a first model and a second model are trained offline. The method comprises the steps of training a first model through a reinforcement learning method according to user information, commodity information and commodity sequence information of historical behaviors of users, and outputting a commodity list of each user. And training the second model by a deep learning method according to the user information and the commodity information bought by the user, and outputting the commodity repurchase cycle of each user. When the online user accesses, filtering the non-repurchase commodities from the commodities to be recommended based on the first model and the second model, and obtaining a recommended commodity list of the current user.

It should be noted that the load balancing may also use a higher-cost hardware load, such as NetScaler, F5, and the like. The secondary cache may also be implemented by other mechanisms, such as Caffeine based on Java 8. Real-time computation may employ a Flink (storm-like real-time data processing framework) equal-flow computation scheme. The DDPG may employ other deep or reinforcement learning algorithms.

The information processing method provided by the embodiment of the disclosure simplifies the marketing production flow of the user based on the scheduling and the resources. And generating advertising creatives for different media according to the recommended real-time use scheduling configuration, so that the method is more automatic. Meanwhile, the real-time calculation of the streaming calculation data, the streaming data and the reinforcement learning algorithm can dynamically adjust the recommendation strategy according to the feedback of the user, meanwhile, the repeated recommendation of commodities in a non-repeated purchasing period is avoided, and the method is more personalized and intelligent. The adoption of Protobuf (data encoding provided by Google) to realize uniform media entry interfacing provides billions of off-site marketing in the prompter periods such as Beijing Dong 618 and Shuangelen.

Fig. 19 schematically shows a block diagram of an information processing apparatus according to an embodiment of the present disclosure.

As shown in fig. 19, the information processing apparatus 1900 may include: an information request receiving module 1910, an information to be shown determining module 1930, and a first sending module 1950.

The information request receiving module 1910 is configured to receive an information request, where the information request includes request body information and transmission object information. The to-be-presented information determining module 1930 is configured to determine, in response to the information request, to-be-presented information corresponding to the sending object information, where the to-be-presented information is for one or more to-be-recommended objects, the presentation information includes a template and element information, the template is determined based on the body information, and the element information is determined based on at least one of the sending object information and the to-be-recommended objects. The first sending module 1950 is configured to feed back the information to be shown to a request body corresponding to the request body information.

In one embodiment, the apparatus 1900 further includes a storage device configured to store a first mapping relationship between the request body information and the information transmission related information before receiving the information request, where the information transmission related information includes the template. Correspondingly, the information to be shown determining module comprises: the device comprises an object to be recommended acquisition submodule, a template determination submodule, an element information acquisition submodule and an assembly submodule. The object to be recommended acquisition submodule is used for acquiring one or more objects to be recommended aiming at the sending object information. The template determining submodule is used for determining a template according to the first mapping relation and the request main body information if the objects to be recommended are obtained and the number of the obtained objects to be recommended is larger than or equal to a number threshold. The element information acquisition submodule is used for acquiring the element information matched with the sending object information and the object to be recommended. The assembling submodule is used for assembling the template and the matched element information to obtain information to be displayed for the object to be recommended.

For example, the object to be recommended acquisition sub-module includes: the device comprises an object to be recommended acquiring unit and an object filtering unit. The object to be recommended acquiring unit is used for acquiring an object to be recommended. The object filtering unit is used for filtering the object to be recommended based on a preset rule to obtain a recommended object. Wherein the preset rule comprises at least one of: and whether the repeated acquisition associated information of the object to be recommended meets the repeated acquisition condition or not. Whether the object to be recommended belongs to a recommendation prohibition object or not, wherein the recommendation prohibition object comprises at least one of the following objects: the sending object forbids the recommending object, the terminal model forbids the recommending object, the request main body forbids the recommending object and the policy forbids the recommending object. Whether the inventory information of the object to be recommended meets the inventory condition or not.

In one embodiment, the to-be-recommended object acquisition unit comprises a historical behavior commodity sequence information acquisition subunit and a first processing unit. The historical behavior commodity sequence information acquiring subunit is used for acquiring the historical behavior commodity sequence information corresponding to the sending object information. The first processing unit is used for processing the sending object information and the historical behavior commodity sequence information by using a first model to obtain information of an object to be recommended.

In particular, the first model comprises a predictor submodel and a parameter update submodel. The input of the predictor model comprises sending object information and historical behavior commodity sequence information, the historical behavior commodity sequence information comprises target marking information, and the output of the predictor model comprises a recommended object and a corresponding accuracy prediction result. The input of the parameter updating submodel comprises an accuracy prediction result and target labeling information corresponding to the sending object information, and the output of the parameter updating submodel comprises an optimization model parameter, wherein the optimization model parameter is obtained based on a first loss function, and the first loss function is constructed based on the accuracy prediction result and the target labeling information corresponding to the sending object information.

Wherein the optimization model parameter corresponds to a direction in which a gradient of the first loss function decreases fastest, and corresponds to the first loss function taking a small value.

In a specific embodiment, the repeatedly acquiring the associated information includes a repeated purchase cycle. The device further comprises: and the second processing unit is used for inputting the sending object information and the information of the object to be recommended into the second model so as to obtain a repurchase period corresponding to the information of the object to be recommended.

In another embodiment, the apparatus 1900 further comprises a second training module for training the second model with training data such that the output of the second model approaches the label information of the training data. The training data comprises sending object information, object information to be recommended, labeling information and at least one of the following: the commodity sequence information of the historical behaviors of the users, the commodity sequence information of the historical behaviors of the users in the same category and the commodity sequence information of the historical behaviors of all the users.

According to an embodiment of the present disclosure, the to-be-shown information determining module 1930 further includes: the hot spot object information acquisition sub-module and the post-processing sub-module. The hot object information obtaining submodule is used for obtaining information to be displayed of the hot object if the information to be displayed of the hot object is not obtained or the number of the objects to be recommended is less than a number threshold, wherein the information to be displayed of the hot object is matched with the information of the sending object. The post-processing submodule is used for taking the information to be displayed of the hot object as the information to be displayed corresponding to the information of the sending object, or complementing the information to be displayed of the object to be recommended by using the information to be displayed of the hot object, so that the quantity of the information to be displayed reaches the quantity threshold value.

In another embodiment, the apparatus 1900 further comprises: the system comprises a second storage module, a scheduling module, an auditing module and a scheduling determining module. Wherein the second storage module is configured to receive and store contract information prior to receiving an information request. And the scheduling module is used for generating scheduling information to be checked of the request main body based on the contract information. The auditing module is used for receiving an auditing result aiming at the scheduling information to be audited, and if the auditing result is passed, the scheduling information to be audited is used as scheduling information. The scheduling determining module is used for matching the request main body information and the request time with the scheduling information after receiving the information request, and responding to the information request if the matching is successful.

Furthermore, the apparatus 1900 may further include: the information sending device comprises an updating operation receiving module and a first updating module, wherein the updating operation receiving module is used for receiving the updating operation after storing a first mapping relation between request main body information and information sending correlation information. The first updating module is used for responding to the updating operation and updating the first mapping relation.

In another embodiment, the apparatus 1900 further comprises: the user operation receiving module and the second updating module. The user operation receiving module is used for receiving the user operation information after the request main body sends the information to be displayed to the main body sending the object information. And the second updating module is used for responding to the user operation information and updating the historical behavior commodity sequence information corresponding to the sent object information so as to at least update the object information to be recommended.

For another example, the apparatus 1900 further includes: and the load distribution module is used for responding to the information request and distributing the load after receiving the information request.

The load shunting module is specifically used for carrying out load shunting by adopting a common component mode. Each node of the server side is provided with an independent load balancer, and the load balancer supports single-node multi-instance deployment so as to distribute loads of multiple instances under the nodes conveniently.

In one embodiment, the apparatus 1900 employs multiple levels of cache, and the element information is stored in the multiple levels of cache. The multi-level cache comprises a server-side cache and a local cache, the server-side cache comprises at least two Redis clusters which are respectively used for storing at least part of the element information, and the server-side cache responds to a local request and transmits cached element information corresponding to the local request to the local cache.

FIG. 20 schematically shows a block diagram of an information handling system according to an embodiment of the disclosure.

As shown in fig. 20, the information processing system is composed of five major subsystems.

And the resource system is used for managing a resource library, and the resource library comprises templates and element information. For example, the system is responsible for basic resource management, and the user inputs related information such as an agent, a channel, media, an advertisement position and the like, and provides data service support for a management system and a playing system.

And the management system is used for scheduling management and determines scheduling information based on the request subject information. For example, the agent is responsible for uploading schedule, service audit and generating corresponding ad placement configuration, such as placement time points, specific media point points, etc., for the play system to use.

And the playing system is used for providing a uniform interface for the outside, receiving the information request, acquiring the template and the element information when the scheduling information meets the scheduling condition, and assembling the template and the element information into information to be displayed so as to be conveniently sent to the request main body corresponding to the request main body information. For example, a unified interface is provided for the outside, a media request is received, the user equipment is switched into an algorithm module, and recommended materials are obtained; and splicing in real time according to the management configuration and the materials, producing the advertisement originality and returning to the media for showing.

And the streaming algorithm system is used for determining the information of the object to be recommended and the element and sending the information to the playing system. For example, on one hand, the historical behaviors of the user are utilized, the latest behaviors of the user are collected in real time, and the latest concerned commodities are calculated in real time by adopting the latest deep learning and reinforcement learning; and on the other hand, the playing system transmits the information of the user equipment, and then recommends the commodities and the related picture information in real time for the playing system to display.

And the real-time computing system is used for updating the resource library and the model parameters based on the acquired user operation. For example, the behavior that results after the media presentation and the user clicks on the advertisement into the Jingdong: browsing, ordering and the like are transmitted into a Storm cluster calculated in real time by using messages, and the Storm cluster is displayed in real time after calculation and statistics are carried out and is used for tracking trends; and moreover, an algorithm system is provided for updating data in real time and modifying model parameters so as to calculate the latest user attention.

Fig. 21 schematically shows a block diagram of an information processing apparatus according to another embodiment of the present disclosure.

As shown in fig. 21, the information processing apparatus 2100 includes: an information request sending module 2110, an information receiving module to be shown 2130 and a second sending module 2150.

The information request sending module 2110 is configured to send an information request to the server, where the information request includes request subject information and sending object information. The information to be shown receiving module 2130 is configured to receive information to be shown for the sending object from a server, where the information to be shown is for one or more objects to be recommended, the display information includes a template and element information, the template is determined based on the subject information, and the element information is determined based on at least one of the sending object information and the objects to be recommended. The second sending module 2150 is configured to send the information to be displayed to the client that sends the object, so that the client that sends the object displays the information to be displayed.

In another embodiment, the apparatus 2100 may further include a user operation information receiving module and a user operation information transmitting module. The user operation information receiving module is used for receiving the user operation information after the information to be displayed is sent to the client side of the sending object. The user operation information sending module is used for sending the user operation information to the server side so that the server side can update the historical behavior commodity sequence information corresponding to the sending object information and the information of the object to be recommended.

Any number of modules, sub-modules, units, sub-units, or at least part of the functionality of any number thereof according to embodiments of the present disclosure may be implemented in one module. Any one or more of the modules, sub-modules, units, and sub-units according to the embodiments of the present disclosure may be implemented by being split into a plurality of modules. Any one or more of the modules, sub-modules, units, sub-units according to embodiments of the present disclosure may be implemented at least in part as a hardware circuit, such as a Field Programmable Gate Array (FPGA), a Programmable Logic Array (PLA), a system on a chip, a system on a substrate, a system on a package, an Application Specific Integrated Circuit (ASIC), or may be implemented in any other reasonable manner of hardware or firmware by integrating or packaging a circuit, or in any one of or a suitable combination of software, hardware, and firmware implementations. Alternatively, one or more of the modules, sub-modules, units, sub-units according to embodiments of the disclosure may be at least partially implemented as a computer program module, which when executed may perform the corresponding functions.

For example, any plurality of the information request receiving module 1910, the to-be-presented information determining module 1930, and the first transmitting module 1950 may be combined into one module to be implemented, or any one of the modules may be split into a plurality of modules. Alternatively, at least part of the functionality of one or more of these modules may be combined with at least part of the functionality of the other modules and implemented in one module. According to an embodiment of the present disclosure, at least one of the information request receiving module 1910, the information to be presented determining module 1930, and the first sending module 1950 may be implemented at least partially as a hardware circuit, such as a Field Programmable Gate Array (FPGA), a Programmable Logic Array (PLA), a system on a chip, a system on a substrate, a system on a package, an Application Specific Integrated Circuit (ASIC), or may be implemented in hardware or firmware by any other reasonable manner of integrating or packaging a circuit, or may be implemented in any one of three implementations of software, hardware, and firmware, or in a suitable combination of any of them. Alternatively, at least one of the information request receiving module 1910, the information to be presented determining module 1930 and the first transmitting module 1950 may be at least partially implemented as computer program modules that, when executed, may perform corresponding functions.

FIG. 22 schematically shows a block diagram of an electronic device according to an embodiment of the disclosure. The electronic device shown in fig. 22 is only an example, and should not bring any limitation to the functions and the scope of use of the embodiments of the present disclosure.

As shown in fig. 22, a computer system 2200 according to an embodiment of the present disclosure includes a processor 2201, which can perform various appropriate actions and processes according to a program stored in a Read Only Memory (ROM)2202 or a program loaded from a storage portion 2208 into a Random Access Memory (RAM) 2203. The processor 2201 may comprise, for example, a general purpose microprocessor (e.g., a CPU), an instruction set processor and/or associated chipset, and/or a special purpose microprocessor (e.g., an Application Specific Integrated Circuit (ASIC)), among others. The processor 2201 may also include on-board memory for caching purposes. The processor 2201 may include a single processing unit or multiple processing units for performing different actions of a method flow according to embodiments of the disclosure.

In the RAM 2203, various programs and data necessary for the operation of the system 2200 are stored. The processor 2201, ROM 2202, and RAM 2203 are connected to each other by a bus 2204. The processor 2201 performs various operations of the method flows according to the embodiments of the present disclosure by executing programs in the ROM 2202 and/or the RAM 2203. Note that the programs may also be stored in one or more memories other than the ROM 2202 and the RAM 2203. The processor 2201 may also perform various operations of method flows according to embodiments of the present disclosure by executing programs stored in the one or more memories.

According to an embodiment of the present disclosure, system 2200 may also include an input/output (I/O) interface 2205, input/output (I/O) interface 2205 also connected to bus 2204. The system 2200 may also include one or more of the following components connected to the I/O interface 2205: an input portion 2206 including a keyboard, a mouse, and the like; an output portion 2207 including a Cathode Ray Tube (CRT), a Liquid Crystal Display (LCD), and the like, a speaker, and the like; a storage portion 2208 including a hard disk and the like; and a communication section 2209 including a network interface card such as a LAN card, a modem, or the like. The communication section 2209 performs communication processing via a network such as the internet. The drive 2210 is also connected to the I/O interface 2205 as needed. A removable medium 2211, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, or the like, is mounted on the drive 2210 as necessary, so that a computer program read out therefrom is installed into the storage section 2208 as necessary.

According to embodiments of the present disclosure, method flows according to embodiments of the present disclosure may be implemented as computer software programs. For example, embodiments of the present disclosure include a computer program product comprising a computer program embodied on a computer readable storage medium, the computer program containing program code for performing the method illustrated by the flow chart. In such an embodiment, the computer program may be downloaded and installed from a network through the communication section 2209, and/or installed from the removable medium 2211. The computer program, when executed by the processor 2201, performs the above-described functions defined in the system of the embodiment of the present disclosure. The systems, devices, apparatuses, modules, units, etc. described above may be implemented by computer program modules according to embodiments of the present disclosure.

The present disclosure also provides a computer-readable storage medium, which may be contained in the apparatus/device/system described in the above embodiments; or may exist separately and not be assembled into the device/apparatus/system. The computer-readable storage medium carries one or more programs which, when executed, implement the method according to an embodiment of the disclosure.

According to embodiments of the present disclosure, the computer-readable storage medium may be a non-volatile computer-readable storage medium, which may include, for example but is not limited to: a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the present disclosure, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. For example, according to embodiments of the present disclosure, a computer-readable storage medium may include ROM 2202 and/or RAM 2203 and/or one or more memories other than ROM 2202 and RAM 2203 described above.

The flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams or flowchart illustration, and combinations of blocks in the block diagrams or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.

Those skilled in the art will appreciate that various combinations and/or combinations of features recited in the various embodiments and/or claims of the present disclosure can be made, even if such combinations or combinations are not expressly recited in the present disclosure. In particular, various combinations and/or combinations of the features recited in the various embodiments and/or claims of the present disclosure may be made without departing from the spirit or teaching of the present disclosure. All such combinations and/or associations are within the scope of the present disclosure.

The embodiments of the present disclosure have been described above. However, these examples are for illustrative purposes only and are not intended to limit the scope of the present disclosure. Although the embodiments are described separately above, this does not mean that the measures in the embodiments cannot be used in advantageous combination. The scope of the disclosure is defined by the appended claims and equivalents thereof. Various alternatives and modifications can be devised by those skilled in the art without departing from the scope of the present disclosure, and such alternatives and modifications are intended to be within the scope of the present disclosure.

Claims

1. An information processing method executed by a server side includes:

receiving an information request, wherein the information request comprises request main body information and sending object information;

in response to the information request, determining information to be shown corresponding to the sending object information, wherein the information to be shown is specific to one or more objects to be recommended, the information to be shown comprises a template and element information, the template is determined based on the main body information, and the element information is determined based on at least one of the sending object information and the objects to be recommended; and

and feeding back the information to be displayed to a request main body corresponding to the request main body information.

2. The method of claim 1, further comprising: storing a first mapping relation between request main body information and information sending associated information before receiving an information request, wherein the information sending associated information comprises a template;

the determining the information to be displayed corresponding to the sending object information comprises:

acquiring one or more objects to be recommended aiming at the sending object information;

if the objects to be recommended are obtained and the number of the obtained objects to be recommended is larger than or equal to a number threshold value, determining a template according to the first mapping relation and the request subject information;

acquiring element information matched with the sending object information and the object to be recommended; and

and assembling the template and the matched element information to obtain the information to be displayed for the object to be recommended.

3. The method of claim 2, wherein the obtaining one or more objects to be recommended for the transmission object information comprises:

obtaining an object to be recommended; and

filtering the object to be recommended based on a preset rule to obtain a recommended object;

wherein the preset rule comprises at least one of:

whether the repeated acquisition associated information of the object to be recommended meets the repeated acquisition condition or not,

whether the object to be recommended belongs to a recommendation prohibition object or not, wherein the recommendation prohibition object comprises at least one of the following objects: a sending object prohibited recommending object, a terminal model prohibited recommending object, a request subject prohibited recommending object and a policy prohibited recommending object,

whether the inventory information of the object to be recommended meets the inventory condition or not.

4. The method of claim 3, wherein the obtaining the object to be recommended comprises:

acquiring historical behavior commodity sequence information corresponding to the sending object information; and

processing the sending object information and the historical behavior commodity sequence information by using a first model to obtain object information to be recommended;

wherein the first model comprises a predictor sub-model and a parameter update sub-model;

the input of the predictor model comprises sending object information and historical behavior commodity sequence information, the historical behavior commodity sequence information has target marking information, the output of the predictor model comprises a recommended object and a corresponding accuracy prediction result, an

The input of the parameter updating submodel comprises an accuracy prediction result and target labeling information corresponding to the sending object information, and the output of the parameter updating submodel comprises an optimization model parameter, wherein the optimization model parameter is obtained based on a first loss function, and the first loss function is constructed based on the accuracy prediction result and the target labeling information corresponding to the sending object information.

5. The method of claim 3, wherein:

the repeatedly acquired associated information comprises a repeated purchase period;

the method further comprises the following steps:

inputting the information of the sending object and the information of the object to be recommended into a second model so as to obtain a repurchase period corresponding to the information of the object to be recommended; wherein the content of the first and second substances,

the second model is trained by:

training the second model with training data such that an output of the second model approaches labeling information of the training data;

the training data comprises sending object information, object information to be recommended, labeling information and at least one of the following: the commodity sequence information of the historical behaviors of the users, the commodity sequence information of the historical behaviors of the users in the same category and the commodity sequence information of the historical behaviors of all the users.

6. The method of claim 2, wherein:

the method further comprises the following steps:

if the objects to be recommended are not obtained or the number of the objects to be recommended is less than the number threshold, obtaining information to be displayed of the hot object, wherein the information to be displayed of the hot object is matched with the information of the sending object; and

taking the information to be displayed of the hot object as the information to be displayed corresponding to the information of the sending object, or complementing the information to be displayed of the object to be recommended by using the information to be displayed of the hot object so as to enable the number of the information to be displayed to reach the number threshold;

and/or

The method further comprises the following steps: after storing the first mapping relationship between the request body information and the information transmission association information,

receiving an updating operation; and

updating the first mapping relationship in response to the updating operation.

7. The method of claim 1, wherein:

the method further comprises the following steps: prior to the receipt of the request for information,

receiving and storing contract information;

generating scheduling information to be checked of the request main body based on the contract information;

receiving an auditing result aiming at the scheduling information to be audited, and if the auditing result is that the scheduling information to be audited is passed, taking the scheduling information to be audited as scheduling information;

and/or

The method further comprises the following steps: after the request for information has been received,

matching the request main body information and the request time with the scheduling information, and responding to the information request if the matching is successful;

and/or

The method further comprises the following steps: after the requesting body transmits the information to be presented to the body transmitting the object information,

receiving user operation information; and

responding to the user operation information, updating historical behavior commodity sequence information corresponding to the sending object information so as to at least update the information of the object to be recommended;

and/or

The method further comprises the following steps: after receiving an information request, responding to the information request, and performing load distribution, wherein the performing of load distribution comprises performing load distribution in a common component mode, each node of a server end is provided with an independent load balancer, and the load balancers support single-node multi-instance deployment so as to perform load distribution on multiple instances under the nodes;

the element information adopts a multi-level cache, the multi-level cache comprises a server-side cache and a local cache, the server-side cache comprises at least two Redis clusters which are respectively used for storing at least part of the element information, and the server-side cache responds to a local request and transmits the cached element information corresponding to the local request to the local cache.

8. An information processing method executed by a media terminal comprises the following steps:

sending an information request to the server, wherein the information request comprises request main body information and sending object information;

receiving information to be displayed aiming at the sending object from a server, wherein the information to be displayed aims at one or more objects to be recommended, the display information comprises a template and element information, the template is determined based on the main body information, and the element information is determined based on at least one of the sending object information and the objects to be recommended; and

and sending the information to be displayed to the client side of the sending object so as to facilitate the client side of the sending object to display the information to be displayed.

9. An information processing apparatus comprising:

the information request receiving module is used for receiving an information request, wherein the information request comprises request main body information and sending object information;

a to-be-displayed information determining module, configured to determine, in response to the information request, to-be-displayed information corresponding to the sending object information, where the to-be-displayed information is for one or more to-be-recommended objects, the display information includes a template and element information, the template is determined based on the subject information, and the element information is determined based on at least one of the sending object information and the to-be-recommended objects; and

and the first sending module is used for feeding back the information to be displayed to the request main body corresponding to the request main body information.

10. An information processing apparatus comprising:

the information request sending module is used for sending an information request to the server side, wherein the information request comprises request main body information and sending object information;

the information receiving module to be displayed is used for receiving information to be displayed aiming at the sending object from a server, wherein the information to be displayed aims at one or more objects to be recommended, the display information comprises a template and element information, the template is determined based on the main body information, and the element information is determined based on at least one of the sending object information and the objects to be recommended; and

and the second sending module is used for sending the information to be displayed to the client side of the sending object so as to facilitate the client side of the sending object to display the information to be displayed.

11. An electronic device, comprising:

one or more processors;

a storage device for storing executable instructions which, when executed by the processor, implement the method of any one of claims 1 to 8.

12. A computer readable storage medium having stored thereon executable instructions which, when executed by a processor, implement a method according to any one of claims 1 to 8.