CN112507218A

CN112507218A - Business object recommendation method and device, electronic equipment and storage medium

Info

Publication number: CN112507218A
Application number: CN202011413027.5A
Authority: CN
Inventors: 曾仰鑫; 叶键晖; 栗金海
Original assignee: Guangzhou Huaduo Network Technology Co Ltd
Current assignee: Guangzhou Huaduo Network Technology Co Ltd
Priority date: 2020-12-03
Filing date: 2020-12-03
Publication date: 2021-03-16

Abstract

The application discloses a business object recommendation method, a business object recommendation device, electronic equipment and a storage medium, wherein the method comprises the following steps: acquiring at least one candidate service object under a target service, wherein each candidate service object corresponds to at least one target label; acquiring at least one reference label generated in a reference service by a service user of the target service, wherein the reference service is different from the target service, and the reference label is determined according to the concerned behavior of the service user in the reference service; determining a recommendation score for each of the candidate business objects based on the at least one reference tag and the at least one target tag; and according to the recommendation score, determining a target business object from the at least one candidate business object and recommending the target business object to the business user. The method and the device can realize accurate service object recommendation.

Description

Business object recommendation method and device, electronic equipment and storage medium

Technical Field

The present application relates to the field of information processing technologies, and in particular, to a method and an apparatus for recommending a business object, an electronic device, and a storage medium.

Background

With the development of internet technology, more and more types of services are provided by a service system, wherein a focus-type service is a common service type, and can provide a content focus function for a user, so that the user can browse interested contents again by using the function. At present, the service system may also recommend related content to the user by using the function, so as to perform targeted content recommendation. However, the current recommendation method is single, and the problem of low accuracy still exists.

Disclosure of Invention

The embodiment of the application provides a business object recommendation method and device, electronic equipment and a storage medium, and the accuracy of business object recommendation can be improved.

In a first aspect, an embodiment of the present application provides a method for recommending a service object, where the method includes: acquiring at least one candidate service object under a target service, wherein each candidate service object corresponds to at least one target label; acquiring at least one reference label generated in a reference service by a service user of the target service, wherein the reference service is different from the target service, and the reference label is determined according to the concerned behavior of the service user in the reference service; determining a recommendation score for each of the candidate business objects based on the at least one reference tag and the at least one target tag; and according to the recommendation score, determining a target business object from the at least one candidate business object and recommending the target business object to the business user.

In a second aspect, an embodiment of the present application provides a service object recommendation device, where the device includes: the candidate object acquisition module is used for acquiring at least one candidate service object under a target service, and each candidate service object corresponds to at least one target label; a reference tag obtaining module, configured to obtain at least one reference tag generated in a reference service by a service user of the target service, where the reference service is different from the target service, and the reference tag is determined according to a behavior of the service user concerning the reference service; a recommendation score evaluation module for determining a recommendation score for each of the candidate business objects according to the at least one reference tag and the at least one target tag; and the target object recommending module is used for determining a target service object from the at least one candidate service object according to the recommending score and recommending the target service object to the service user.

In a third aspect, an embodiment of the present application provides an electronic device, including: a memory; one or more processors coupled with the memory; one or more application programs, wherein the one or more application programs are stored in the memory and configured to be executed by the one or more processors, and the one or more programs are configured to perform the business object recommendation method provided by the first aspect.

In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, where a program code is stored in the computer-readable storage medium, and the program code may be called by a processor to execute the service object recommendation method provided in the first aspect.

According to the service object recommendation method, the device, the electronic equipment and the storage medium provided by the embodiment of the application, at least one candidate service object under a target service is obtained, wherein each candidate service object corresponds to at least one target label, at least one reference label generated in the reference service by a service user of the target service is obtained, the reference service is different from the target service, the reference label is determined according to the attention behavior of the service user in the reference service, then the recommendation score of each candidate service object is determined according to the at least one reference label and the at least one target label, and the target service object recommended to the service user is determined and recommended from the at least one candidate service object according to the recommendation score. According to the method and the device, the attention behaviors of the user under different service types are subjected to labeling processing, and the interest of the user under different service types can be obtained in a classifying mode, so that when the service object under the target service is recommended to the service user under the target service, the acceptance degree of the service user to each candidate service object can be comprehensively evaluated according to the interest of the service user under other service types, and the accurate recommendation of the service object for the service user can be realized according to the evaluation result.

Drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed to be used in the description of the embodiments are briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present application, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.

Fig. 1 shows an application scenario diagram of a service object recommendation method provided in an embodiment of the present application.

Fig. 2 shows a business process diagram provided in an embodiment of the present application.

Fig. 3 shows an interface diagram of a focused data storage according to an embodiment of the present application.

Fig. 4 shows another schematic business process provided in this embodiment of the present application.

Fig. 5 is a flowchart illustrating a business object recommendation method according to an embodiment of the present application.

Fig. 6 is a flowchart illustrating a business object recommendation method according to another embodiment of the present application.

Fig. 7 is a flowchart illustrating a business object recommendation method according to another embodiment of the present application.

Fig. 8 is a flowchart illustrating a business object recommendation method according to yet another embodiment of the present application.

Fig. 9 is a flowchart illustrating a step S430 in a business object recommendation method according to yet another embodiment of the present application.

Fig. 10 is a flowchart illustrating a step S432 in a business object recommendation method according to yet another embodiment of the present application.

Fig. 11 is a flowchart illustrating a business object recommendation method according to yet another embodiment of the present application.

Fig. 12 shows a block diagram of a business object recommendation device according to an embodiment of the present application.

Fig. 13 shows a block diagram of an electronic device according to an embodiment of the present application.

Fig. 14 illustrates a storage unit for storing or carrying a program code for implementing the business object recommendation method according to the embodiment of the present application.

Detailed Description

In order to make the technical solutions better understood by those skilled in the art, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application.

Referring to fig. 1, fig. 1 is a schematic diagram illustrating an application scenario of a service object recommendation method provided in an embodiment of the present application, where the application scenario includes a service system 10 provided in the embodiment of the present application. The service system 10 includes: a terminal device 100 and a server 200. Wherein, the terminal device 100 and the server 200 are located in a wireless network or a wired network, and the terminal device 100 and the server 200 can perform data interaction. In some embodiments, there may be a plurality of terminal devices 100, the server 200 may be communicatively connected to a plurality of terminal devices 100, a plurality of terminal devices 100 may also be communicatively connected to each other through the internet, and the server 200 may also be used as a transmission medium to implement data interaction with each other through the internet.

In this embodiment, the terminal device 100 may be a mobile phone, a smart phone, a notebook computer, a desktop computer, a tablet computer, a Personal Digital Assistant (PDA), a media player, a smart television, a wearable electronic device, and the like, and a specific type of the terminal device may not be limited in this embodiment. The server 200 may be a single server, or a server cluster, or a local server, or a cloud server, and a specific server type may not be limited in this embodiment of the application.

In some embodiments, a client may be installed in the terminal device 100. The client may be a computer Application (APP) installed on the terminal device 100, or may be a Web client, which may refer to an Application developed based on a Web architecture. When a user logs in through an account at a client, all information corresponding to the account can be stored in the storage space of the server 200. The information corresponding to the account includes information input by the user through the client, information received by the user through the client, and the like.

In some implementations, the client may be an application for providing content services, such as a novel reading platform, an electronic book payment platform, a video platform, a gaming platform, and the like. In some embodiments, the client may provide an attention type service so that a user may utilize the client to generate attention behavior for content of interest. In some embodiments, the business system may store the attention behavior data of the user to the server, and may also send a relevant notification or recommend relevant content to the client used by the user according to the attention behavior data.

For example, when a user pays attention to a novel author on a novel reading APP, after the novel author issues a new novel, the business system may send a notification to the user according to the attention relationship to recommend a new book of the novel author. For another example, if the user "votes" for a chapter, the user votes for a novel story, and the business system may send a story ranking list to the "voted" or voted user, and recommend books listed on the story. For another example, a user collects a certain book on a certain book e-commerce website, and the website can recommend the user when the price of the book changes or the same type of book is favorable.

Referring to fig. 2, fig. 2 is a schematic view illustrating a business process provided by an embodiment of the present application. As can be seen, the specific business processes of the above-mentioned services of the type of interest can be summarized as follows: the user is used as a follower (namely an initiator of a follow-up behavior in a service), and can initiate a request to a service system to establish a follow-up relationship with a certain service object, and the service system can record follow-up relationship data of the user; the concerned person can inquire the concerned list, and can also inquire whether the concerned person has concerned about a certain business object, and can further obtain the content service provided by the concerned object (namely the concerned action receiver in the business) through the concerned list; the concerned object can inquire which users concern themselves (namely, fan list); when the concerned object releases new content or other changes, the business system can inform the corresponding concerned person according to the fan list, and can also recommend other business objects to the concerned person according to the characteristics of the concerned object, so as to provide new content service. The attention type services such as attention, subscription, voting, praise, collection and the like all conform to the flow.

The inventor finds that the development efficiency is low when the attention type services are developed at present through research. Specifically, if each business scenario develops its own relationship storage service, then there will be different implementations of multiple sets of homogeneous services in the same system. For example, an interface for determining whether a user has established an attention relationship, each service needs to be provided separately, and the multiplexing rate is low. And when some operation activities need to be quickly brought online, if the relationship needs to be concerned, the workload of developing the storage service is added to the workload evaluation, so that the development efficiency is greatly reduced. Moreover, because each service scene develops its own relationship storage service, the concerned behavior data between different services is not communicated, so that the recommended data of each service scene cannot refer to the concerned behavior data of other services, and the accuracy of the recommended data is low.

In the embodiment of the present application, after long-term research aiming at the problem of low development efficiency, the inventor provides a general attention system provided in the embodiment of the present application, so that the service system does not need to implement attention data storage by itself, but the general attention system takes over attention data storage uniformly, thereby improving the service development efficiency.

In some embodiments, the unified takeover attention data storage may be storage of forward relation data and reverse relation data. Specifically, for the attention type service, a mapping relationship is actually established between the user and the service object, and from the perspective of the attention person, the mapping relationship is [ attention person- > attention object ], which is forward relationship data; from the perspective of the object of interest, the mapping is [ object of interest- > attendee ], which is the inverse relationship data. Because the logic of the mapping relationship of different attention type services is the same, the relationship data can be taken over and stored by the general attention system of the application. In some embodiments, the forward relationship data and the reverse relationship data may be checked and synchronized periodically to maintain data final consistency.

In other embodiments, the unified takeover concerns data storage, or may be storage of only forward relationship data, that is, the forward relationship data is basic data that must be stored, and the reverse relationship data is data that can be optionally stored. It is to be understood that the inverse relationship data is stored, primarily for convenience in retrieving a list of followers of the object of interest. However, since some service scenarios do not need to provide the attendee list, and some service scenarios need to provide, it may be determined whether to store the reverse relationship data according to the actual service scenario, which is not limited herein. For example, some business scenarios may have some hot business objects, so that the data volume of the attendee list thereof is increased sharply, thereby causing a large data access pressure, and the data access pressure can be dispersed by providing a reverse relationship data storage for the hot business objects.

Since the general attention system may not need to know the meaning of the attention data in a specific service scene and store the complete service information of the attention data for different attention people or attention objects, in some embodiments, an Identifier (ID) of the attention people or attention objects may be used as a storage basis for different attention people or attention objects when storing the attention data. For example, when storing attention relationship data of a certain user, the storage format of the basic data (i.e., forward relationship data) may be [ attention person identification ID: object of interest identification ID). It is understood that, when the reverse relationship data needs to be stored, the storage format thereof may be [ object identification ID of interest: attendee identification ID).

For example, for a book attention business in a novel reading APP, the attention person is a reading user, and the attention object is a book. When a user pays attention to a novel, the information stored in the general attention system is the user identification ID and the book identification ID, and the complete user information and the complete book information do not need to be stored.

Therefore, even if a plurality of services are simultaneously accessed to the general attention system, the storage data volume can not be increased too fast, the storage cost is reduced, the storage format (the field of the database table) of the bottom data does not need to be adjusted, the complexity of data storage is greatly reduced, and the development cost is also reduced.

In some embodiments, in order to ensure that different services of interest do not affect each other, a unique service identification ID may be assigned to each accessed service of interest. For example, when storing the attention relationship data of a certain user of a service, the basic storage format may be [ service identification ID: the attendee identification ID: object of interest identification ID).

In some embodiments, the storage resource instance implementation may be configurable such that each service may choose to allocate a separate storage database instance. The type of the database may be MySQL/MongoDB/KV, etc., and is not limited herein. The types of the databases corresponding to different services may be the same or different, and are not limited herein.

It can be understood that, after different services are accessed, independent service identification IDs and corresponding storage resources can be allocated, so that even if a certain service system or storage resource fails, normal operation of other services is not affected, and it is sufficient to troubleshoot and recover the problem service.

Referring to fig. 3, fig. 3 is a schematic diagram of an interface for data storage of interest provided by the present application. Specifically, when the service of interest needs to access the universal system of the present application, relevant service information may be filled in to apply for a service identifier and a storage resource, such as a service identifier ID, a database type, and a database name corresponding to different service names shown in fig. 3. Optionally, whether the reverse relationship data is stored or not may be filled, and if yes, the subsequent general attention system may automatically record the reverse relationship data when storing the attention data.

In some embodiments, a general interface may also be provided for the service scenarios of various types of attention, so that various execution logics of the attention relationship may be implemented by calling the general interface, including adding/deleting the attention relationship, querying whether the attention relationship exists, querying a list and a number of the attention objects of a certain attention object, and the like. The basic parameter of the universal interface may be at least one of a service identification ID, an attendee identification ID, and an attendee identification ID. In some embodiments, different execution logic may correspond to different universal interfaces, so that when a target execution logic needs to be implemented, a target universal interface corresponding to the target execution logic may be called.

Specifically, adding/deleting the concern relationship data may be realized by modifying the forward relationship data; inquiring whether the concern relationship exists or not can be realized by inquiring forward relationship data; inquiring a list and the number of the concerned objects of a certain concerned person can be realized by inquiring forward relation data; the query of the list and number of the attendees of a certain attended object can be realized by querying the reverse relationship data.

In some embodiments, if the business system needs to provide services in different data centers (machine rooms), the service instances and the storage resources of the general attention interface can be flexibly deployed nearby, so that the delay of the interface is reduced.

Referring to fig. 4, for example, with the general attention system of the present application, an attention behavior flow may be: after the service system is accessed to the general attention system of the application, an attention person initiates a request, request content is sent to a specific service system, after the service system receives the request, a corresponding general interface of the general attention system can be called according to the request type, the general attention system receives the request, a corresponding storage database instance is found according to parameters (a service identifier, an attention person identifier and an attention person identifier) in the request, data modification or data query is carried out, after the processing is completed, reply content can be sent to the service system, the service system can process the reply content and return a final result to the attention person. In some embodiments, when the reply received by the service system includes the relevant object identifier, the service system may completely fill the identifier information corresponding to the relevant object identifier according to its own needs, so as to display the complete content to the user.

In the embodiment of the present application, in order to solve the problem of low recommendation accuracy, on the basis of the general attention system, the embodiment of the present application further provides a method, an apparatus, a system, an electronic device, and a storage medium for recommending a service object, which can improve the accuracy of recommending a service object. The following will be described in detail by way of specific examples.

Referring to fig. 5, fig. 5 is a schematic flowchart illustrating a flow of a service object recommendation method provided in an embodiment of the present application, and the method may be applied to an electronic device, where the electronic device may be a server corresponding to the general attention system, a server corresponding to the service system, or the terminal device, and is not limited herein. The business object recommendation method can comprise the following steps:

step S110: at least one candidate service object under the target service is obtained, and each candidate service object corresponds to at least one target label.

Because the concerned behavior data among different services is not communicated, the recommended data of a certain service is determined only by the concerned behavior data of the current service. However, the attention behavior of the user in the current service may not completely reflect the actual interest preference of the user, so if the service object is recommended to the user only depending on the attention behavior of the current service, the recommendation accuracy may not be high. Therefore, in the embodiment of the present application, based on the above general attention system, the interest preferences of the user reflected in other services may be determined, and then the interest preferences of the user between different services may be integrated to accurately recommend the service object under the current service to the user.

Wherein, the business object can be an object related to the service content provided by the business. The service is a set of services provided for users, and may be an online video service, an online audio service, an online text service, and the like. The video may be a movie, a television show, or a user-made video, etc. The audio may be a pop song, music, or user-made audio, etc. Text such as a novel, a toolbook, or a professional dictionary, etc. As one way, the business object may be the specific service content provided by the business. E.g. a certain novel, a certain video, a certain music, etc. Alternatively, the business object may be a character related to the service content, which may be a creator of the service content or a participant of the service content. For example, an author of a novel, an originator of a homemade video, a participant or director in a video, a singer of a piece of music, etc. As another way, the service object may also be a service subscriber, that is, a user of the service. For example, in a novel reading APP, a reading user A, a reading user B, and the like. In some embodiments, a service user may also be understood as a terminal used by the user.

In the embodiment of the application, when a service object under the current service needs to be recommended to a user, the electronic device may obtain at least one candidate service object under the target service. The target service may be a service that needs to be recommended by a service object, and the candidate service object is an object related to the service content provided by the target service, which may be a service object that needs to be recommended and evaluated.

In some embodiments, the candidate service object may be a service object having a certain recommendation probability in the target service for the service user, so that the candidate service object may be further recommended and evaluated by combining interest preferences reflected by the user in other services, thereby filtering out a candidate service object with inaccurate recommendation, and implementing accurate recommendation of the service object.

As a mode, the electronic device may obtain a hot service object under the target service, as a candidate service object under the target service. The hotspot business object can be understood as a business object which is concerned or welcomed by the broad masses, and can be determined by the user access times or frequency and the user click times or frequency of the business object. It can be understood that, because most users keep a certain interest and curiosity in the hot spot service object, the acceptance degree of the hot spot service object by the users is generally higher, that is, the hot spot service object has a certain recommendation probability, so that the hot spot service object can be used as a candidate service object for subsequent further promotion and evaluation, and the recommendation accuracy and efficiency of the service object are improved.

As another mode, the electronic device may preliminarily determine a recommended service object according to the attention behavior data generated by the user under the target service, as a candidate service object under the target service. It can be understood that, since the preliminarily determined and recommended service object is the object to be recommended which is determined according to the attention behavior data under the target service, the user has a certain acceptance to the preliminarily determined service object, that is, the preliminarily determined service object has a certain recommendation probability, so that the service object can be used as a candidate service object to perform subsequent further promotion and evaluation to screen out the service object with higher acceptance for recommendation, and the recommendation accuracy of the service object is improved while the recommendation efficiency is also improved.

In some embodiments, the attention behavior of the user under the target service can be tagged, so that the interest preference of the user under the target service type can be classified. As a mode, the target service may preset a corresponding service tag for each service object, and the service tag may embody a style and a type characteristic of the service object, so that when a user generates an attention behavior to the service object, the attention behavior may be marked with the service tag corresponding to the service object. Since the service label pre-established by the target service may be incomplete or incorrect, the service label may be set and modified by the user. The particular label arrangement is not limited herein.

In some embodiments, the service tags correspondingly set for different service types may be the same, may be different, or may be partially the same, which is not limited herein. In the embodiment of the application, when the electronic device acquires at least one candidate business object under the target business, the electronic device may also acquire at least one target label corresponding to each candidate business object to determine the style type characteristics of the candidate business object. The target label may be a service label correspondingly set under the target service, and may be determined correspondingly according to the style type feature of the candidate service object.

It can be understood that, when a candidate service object is a service object having a certain recommendation probability for a service user in a target service, at least one target tag corresponding to the candidate service object may be used to reflect a certain interest preference of the service user in the target service, but may not completely reflect an actual interest preference of the user, so that a further recommendation evaluation needs to be performed on the candidate service object subsequently.

In some embodiments, a tag list corresponding to the attention behavior may also be established in the general attention system, so that the interest preference of the user may be determined according to the tag list. As one way, the tag list may be stored in correspondence with the data of interest. For example, when storing attention relationship data of a certain user of a service, the storage correspondence may be [ service identification ID ] - [ attention person identification ID ] - [ attention object identification ID ] - [ tag 1, tag 2, tag 3, … …, tag n ].

Step S120: and acquiring at least one reference label generated in a reference service by a service user of the target service, wherein the reference service is different from the target service, and the reference label is determined according to the concerned behavior of the service user in the reference service.

In the embodiment of the application, the electronic device may acquire at least one reference tag generated in the reference service by a service user of the target service to determine the interest preference reflected in the reference service by the user. The reference service may be a service for providing a recommendation basis for a candidate service object of a target service, and the reference service is different from the target service. In some embodiments, the reference service may be one or more, and is not limited herein.

It can be understood that, since the above-mentioned general attention system uniformly stores and processes attention data of different services, it is equivalent to get through attention behavior data between different services. Therefore, the electronic device can acquire the attention behavior data of the user in different services according to the general attention system, so that the service tags generated by the user in different services can be determined according to the attention behavior data, and the interest preference of the user in different services can be further determined according to the service tags.

The service user of the target service may be understood as a user using the target service, and the reference label may be a service label correspondingly set in the reference service, and may be determined correspondingly according to the style and type characteristics of the service object concerned by the service user in the reference service. As a mode, the electronic device may obtain attention behavior data generated by a service user of a target service in a reference service, then determine a corresponding service tag list according to the attention behavior data, and obtain at least one reference tag generated by the service user in the reference service according to the service tag list. That is, the service user of the target service uses not only the target service but also the reference service.

In some embodiments, the reference traffic may be of the same or similar traffic class as the target traffic. For example, when the target service is a pay-for-reading service of an electronic book, the reference service may be a book attention service of a novel, and is also a text service category. In other embodiments, the reference service may also be a service class that is different or dissimilar from the target service. For example, when the target service is a video service, the reference service may be a book attention service of a novel, one is a video service category, and the other is a text service category. The relationship between the specific reference service and the target service is not limited herein.

Step S130: and determining a recommendation score of each candidate business object according to the at least one reference label and the at least one target label.

In this embodiment of the application, the electronic device may determine, according to the obtained at least one reference tag and the obtained at least one target tag, a recommendation score of each candidate business object in the at least one candidate business object. Therefore, the receiving degree of each candidate service object under the target service is comprehensively evaluated according to the interest preference of the user under other service types, and accurate recommendation can be realized according to the evaluation result subsequently.

It can be understood that, since the at least one reference tag may be used to characterize the interest preference reflected by the user in the reference service, and the at least one target tag may be used to characterize the interest preference reflected by the user in the target service, the recommendation accuracy of each candidate service object may be evaluated by integrating the interest preferences reflected by the user in different services between the reference service and the target service, so as to obtain the recommendation score of each candidate service object, so as to determine whether to recommend the candidate service object according to the recommendation score.

In some embodiments, for each candidate service object, when the interests and hobbies reflected by the user in different services of the reference service and the target service are approximate, the recommendation accuracy of the candidate service object is considered to be higher, the probability of the candidate service object being accepted by the user is considered to be higher, and therefore the determined recommendation score of the candidate service object is higher. Similarly, when the difference between the interests and the hobbies reflected by the user in the different services of the reference service and the target service is large, the recommendation accuracy of the candidate service object is considered to be low, the probability of the user receiving the candidate service object is considered to be low, and therefore the recommendation score of the candidate service object is determined to be low.

In some embodiments, the electronic device may determine, for each candidate business object, at least one target tag corresponding to the candidate business object, and then determine, according to the matching degree of at least one reference tag and the at least one target tag, a recommendation score of the candidate business object, so as to obtain a recommendation score of each candidate business object. The matching degree of the at least one reference tag and the at least one target tag may be an importance matching degree of the tags, or a similarity matching degree of the tags, which is not limited herein. For example, the importance of the tag and the degree of matching of the similarity may be combined.

In some embodiments, the higher the recommendation score of the candidate business object, the higher the probability that the candidate business object is accepted by the business user, or the lower the recommendation score of the candidate business object, the higher the probability that the candidate business object is accepted by the business user. The setting is not limited herein, and may be set as appropriate according to actual needs.

Step S140: and according to the recommendation score, determining a target business object from the at least one candidate business object and recommending the target business object to the business user.

In the embodiment of the application, when the recommendation score of each candidate service object is obtained, a target service object recommended to a service user can be determined from at least one candidate service object according to the recommendation score, and the target service object is recommended. It can be understood that, when the recommendation score of the candidate service object indicates that the probability that the candidate service object is accepted by the service user is relatively high, the candidate service object may be recommended, and when the recommendation score of the candidate service object indicates that the probability that the candidate service object is accepted by the service user is relatively low, the candidate service object may not be recommended.

In some embodiments, at least one candidate service object may be ranked according to the recommendation score of each candidate service object to obtain a ranking result of each candidate service object, and then a target service object may be selected from the at least one candidate service object according to the ranking result to recommend the target service object to a service user. Optionally, the target service object pushed to the service user each time may be different, so as to reduce the user aversion. The at least one candidate service object may be ranked according to a recommendation score from small to large or from large to small, which is not limited herein.

In some embodiments, the candidate business object at the top or the back of the ranking result may be determined as the target business object recommended to the business user. Specifically, when the higher the recommendation score of the candidate service object is, the higher the probability that the candidate service object is accepted by the service user is, the candidate service object with the top ranking result may be determined as the target service object recommended to the service user; when the recommendation score of the candidate service object is smaller and the probability that the candidate service object is accepted by the service user is higher, the candidate service object with the later ranking result can be determined as the target service object recommended to the service user. The setting is not limited herein, and may be set as appropriate according to actual needs.

In other embodiments, the candidate business object with the highest or lowest recommendation score may be determined as the target business object recommended to the business user. Specifically, when the higher the recommendation score of the candidate service object is, the higher the probability that the candidate service object is accepted by the service user is, the candidate service object with the highest recommendation score may be determined as the target service object recommended to the service user; when the recommendation score of the candidate service object is smaller and the probability that the candidate service object is accepted by the service user is higher, the candidate service object with the lowest recommendation score can be determined as the target service object recommended to the service user. The setting is not limited herein, and may be set as appropriate according to actual needs.

In some embodiments, a plurality of services may also apply for establishing an association relationship, so that when recommending content to a service user under a target service, the association service of the target service may be directly used as the reference service according to the association relationship, so as to implement content recommendation with reference to the attention data of the association service. Specifically, when the service applies for establishing the association relationship, the service object list of each service and the type of the service object may be reported. The service system can establish preference label data of the user according to the concerned behavior data of the user in the service. When the business party requests to recommend data for the user, the business system can calculate according to the preference tag data of the user and recommend related types of articles. Therefore, after more associated services are accessed to the system, more and more comprehensive hobby labels can be added to the user, and recommended articles can be expanded to a larger range.

For example, a novel website (mark 1001), a video website (mark 1002), and a game platform (mark 1003) are added to the general attention system as business parties. User a is at a novel website, paying attention to a set of novels and authors. The video website service requests recommendation data, and the system can recommend content such as videos (interview records, TV shows, movies, shows) adapted by novels that have focused on the writer, or videos of the same type. The game platform service requests recommendation data, and the system can recommend the contents of games (web games, hand games) and the like which are adapted by novels of the concerned writer.

The service object recommendation method provided by the embodiment of the application obtains at least one candidate service object under a target service, wherein each candidate service object corresponds to at least one target tag, and obtains at least one reference tag generated by a service user of the target service in a reference service, wherein the reference service is different from the target service, the reference tag is determined according to the attention behavior of the service user in the reference service, and then the recommendation score of each candidate service object is determined according to the at least one reference tag and the at least one target tag, so that the target service object recommended to the service user is determined and recommended from the at least one candidate service object according to the recommendation score. According to the method and the device, the attention behaviors of the user under different service types are subjected to labeling processing, and the interest of the user under different service types can be obtained in a classifying mode, so that when the service user under the target service recommends the service object under the target service, the acceptance degree of the service user to each candidate service object can be comprehensively evaluated according to the interest of the service user under other service types, and the accurate recommendation of the service object for the service user can be realized according to the evaluation result.

Referring to fig. 6, fig. 6 is a schematic flowchart illustrating a business object recommendation method according to another embodiment of the present application, which is applicable to an electronic device, and the business object recommendation method may include:

step S210: at least one candidate service object under the target service is obtained, and each candidate service object corresponds to at least one target label.

Step S220: and acquiring at least one reference label generated in the reference service by the service user of the target service.

In the embodiment of the present application, step S210 and step S210 may refer to the foregoing embodiments, and are not described herein again.

Step S230: for each of the target tags, a designated reference tag that matches the target tag is determined from the at least one reference tag.

In some embodiments, after obtaining the at least one target tag and the at least one reference tag, the electronic device may determine, for each target tag, a specific reference tag matching the target tag from the at least one reference tag, so as to determine whether the target tag of the target service appears in a reference tag generated by a user in the reference service.

It will be appreciated that since different services may have different service labels, different service labels may exist in the target label of the target service and the reference label under the reference service. When there is a specific reference tag matching the target tag from among the at least one reference tag, it may be considered that the user has the same or similar interest and taste in different services of the reference service and the target service, and thus, the recommendation score of each candidate service object may be determined according to the importance degree of the same or similar interest and taste in different services. The designated reference tag matching with the target tag may be the same or similar to the target tag.

Specifically, the electronic device may determine, for each candidate business object, at least one target tag corresponding to the candidate business object, and determine, for each target tag in the at least one target tag, a designated reference tag matched with the target tag from the at least one reference tag, so as to obtain the designated reference tag corresponding to each candidate business object. Wherein, the number of the assigned reference labels corresponding to each candidate service object can be one, which indicates that the user has a same or similar interest in the different services of the reference service and the target service; or a plurality of interest preferences may be present in different services of the reference service and the target service, which are not limited herein.

Step S240: and acquiring the importance score of the specified reference label in the reference service as a first importance score of the target label in the reference service.

In some embodiments, after obtaining the designated reference tag matching the target tag, the importance score of the designated reference tag in the reference service may be obtained as the first importance score of the target tag in the reference service. Therefore, the importance degree of the common interests and hobbies of the reference service and the target service in the reference service is obtained.

In a service, the importance score of each service tag generated by the attention behavior of a certain user may be obtained by first obtaining all tag lists corresponding to all attention behaviors generated by the user in the service, and then obtaining the number of tags in all tag lists as the total number of reference tags in the service for the user. Then, the occurrence number of each service tag, that is, the total number of each service tag, can be determined according to all the tag lists. The ratio of the total number of each service label to the total number of the reference labels of the user in the service is obtained, so as to obtain the proportion of the total number of the reference labels of each service label in the service, and then the ratio of each service label can be used as the importance score of each service label in the service. That is, the importance score of a certain service tag is the number of occurrences of a certain service tag/the number of occurrences of all tags.

For example, the user generates 2 service tags in the novel interest service: the city says the emotion and the immortal man trusts. In the business, the total number of times of tagging the business object by the user's total attention behaviors is 10000, wherein the number of times of tagging the "urban sentiment" label is 5000, the number of times of tagging the "swordsman truer" label is 1000, the importance score of the "urban sentiment" label is 0.50(5000/10000), and the importance score of the "swordsman truer" label is 0.10 (1000/10000).

Specifically, all the tag lists corresponding to all the attention behaviors generated by the service user in the reference service may be acquired, and then the number of tags in all the tag lists is acquired as the total number of reference tags in the reference service by the service user. Then, the electronic device may determine the number of occurrences of the designated reference tag, that is, the total number of the designated reference tags, according to all the tag lists. And obtaining the ratio of the total number of the appointed reference tags to the total number of the reference tags in the reference service of the service user as the importance score of the appointed reference tags in the reference service.

In some embodiments, for each target tag, when there is no designated reference tag matching the target tag in at least one reference tag, it may be considered that the target tag of the target service does not appear in the reference tags generated in the reference service by the user, and at this time, a preset default importance score may also be obtained as the first importance score of the target tag in the reference service. The default importance score may be set as appropriate according to the specific situation, and may be set to 0.01, for example. Therefore, importance score evaluation can be carried out on each target label of the candidate business object, the influence on the accuracy of subsequent recommendation evaluation due to data loss is avoided, and the accuracy of business object recommendation is improved.

Step S250: determining a second importance score of the target tag in the target business.

In some embodiments, the electronic device may obtain a second importance score of the target tag in the target service to obtain an importance degree of interest and hobbies shared by the reference service and the target service in the target service, so that whether the recommendation of the candidate service object is accurate may be determined according to the importance degree of the interest and hobbies in different services.

Specifically, all the tag lists corresponding to all the attention behaviors generated by the service user in the target service may be acquired, and then the number of tags in all the tag lists is acquired as the total number of reference tags in the target service by the service user. Then, the electronic device may determine the number of occurrences of the target tag, that is, the total number of the target tags, according to all the tag lists. And obtaining the ratio of the total number of the target labels to the total number of the reference labels of the service users in the target service as a second importance score of the target labels in the target service.

Step S260: determining a recommendation score for each of the candidate business objects based on the first importance score and the second importance score.

In some embodiments, the recommendation score for each candidate business object may be determined based on a magnitude of a difference between the first importance score and the second importance score. Specifically, the electronic device may determine, for each candidate business object, at least one target tag corresponding to the candidate business object, so as to determine, for each target tag of the at least one target tag, a difference between a first importance score of each target tag in a reference business and a second importance score of each target tag in a target business, so as to obtain a difference in importance degree of different interest preferences of the candidate business object in different businesses, so that a recommendation score of each candidate business object may be determined.

As one approach, the recommendation score for each candidate business object may be determined based on the difference between the first importance score and the second importance score. In order to make the apparent difference more obvious, as another way, the recommendation score of each candidate business object can be determined according to the power of the difference value of the first importance score and the second importance score. The power square may be a 2 nd power, i.e., a square.

In some embodiments, when the difference between the first importance score of each target tag at the reference service and the second importance score at the target service is relatively large, the determined recommendation score of the candidate service object may be relatively large, and when the difference between the first importance score of each target tag at the reference service and the second importance score at the target service is relatively small, the determined recommendation score of the candidate service object may be relatively small. The method is not limited and can be set reasonably according to specific conditions.

It can be understood that, when the difference between the importance scores of the target tag in the reference service and the target service is relatively large, it can be considered that the importance degrees of the target tag in different services are different, that is, when the style type reflected by the target tag is in different services, the degree of the interest and preference of the user is different, and the probability that the candidate service object corresponding to the target tag is accepted by the user is uncertain, so that the recommendation priority of the candidate service object can be put back to correspond to a lower recommendation priority, that is, when the recommendation score of the candidate service object is relatively large, the recommendation priority is lower.

On the contrary, when the difference between the first importance score of the reference service and the second importance score of the target service is smaller, the importance degrees of the target tags in different services can be considered to be the same or similar, that is, when the style types reflected by the target tags are in different services, the interest and preference degrees of the users are the same, the candidate service object determined by the target service can be considered to substantially reflect the actual interest and preference of the users, and the candidate service object corresponding to the target tag is recommended accurately, so that the recommendation priority of the candidate service object can be put forward to correspond to a higher recommendation priority. That is, the recommendation priority is higher when the recommendation score of the candidate business object is smaller.

For example, suppose reference service a is a novel book attention service, and User1 is concerned with book1, whose reference labels are "campus love", "delay beauty". The target service B is an electronic book paying reading service, a User1 comes to the service B, and candidate electronic books recommended to the User by the service B are respectively:

the book recom _ book _ aa corresponds to a target label of 'campus love', 'delay beauty';

and the book recom _ book _ bb is corresponding to the target label of the book called 'eastern hallucinations'.

Assuming that in reference service a, the importance score of the "campus love" tag is 0.3, the importance score of the "delay" tag is 0.1, the "eastern fantasy" tag User1 was not generated (absent) in service a, a default importance score of 0.01 was used; in the target service B, the importance score of the "campus love" tag is 0.25, the importance score of the "delay beauty" tag is 0.15, and the importance score of the "east fantasy" tag is 0.1.

The electronic device may obtain the recommendation score of the candidate book recom _ book _ aa as follows: (0.3-0.25)²+(0.1-0.15)²The recommendation score of the candidate book recom _ book _ bb is 0.005: (0.1-0.01)²0.0081. It can be seen that the recommendation score of the candidate book recom _ book _ aa is relatively small, the difference of the importance degree of the service tag of the candidate book recom _ book _ aa in different services is considered to be relatively small, the actual interest preference of the user is basically reflected, and the probability of being accepted by the user is considered to be relatively high. The recommendation score of the candidate book recom _ book _ bb is relatively large, so that the difference between the importance degrees of the service tags of the candidate book recom _ book _ bb in different services is relatively large, the book may not reflect the actual interest preference of the user, and the probability of being accepted by the user is considered to be small. So that the candidate book recom _ book _ aa can be subsequently preferentially selected for recommendation to the User 1.

Step S270: and according to the recommendation score, determining a target business object from the at least one candidate business object and recommending the target business object to the business user.

In the embodiment of the present application, step S270 can refer to the foregoing embodiments, and is not described herein again.

The business object recommendation method provided by the embodiment of the application, by obtaining at least one candidate business object under a target business, wherein each candidate business object corresponds to at least one target label, and obtaining at least one reference label generated in a reference business by a business user of the target business, wherein the reference business is different from the target business, the reference label is determined according to the attention behavior of the business user in the reference business, then for each target label, a specified reference label matched with the target label is determined from the at least one reference label to obtain the importance score of the specified reference label in the reference business as a first importance score of the target label in the reference business, and determine a second importance score of the target label in the target business according to the first importance score and the second importance score, and determining the recommendation score of each candidate service object, and determining and recommending a target service object recommended to a service user from the at least one candidate service object according to the recommendation score. According to the method and the device, the attention behaviors of the user under different service types are subjected to labeling processing, the importance scores of all labels are generated, and the interest of the user under different service types and the importance of all interest can be obtained through classification, so that when the service user under the target service recommends the service object under the target service, the acceptance degree of the service user to all candidate service objects can be comprehensively evaluated according to the importance of the interest of the service user under other service types, and the accurate recommendation of the service object aiming at the service user can be realized according to the evaluation result.

Referring to fig. 7, fig. 7 is a schematic flowchart illustrating a business object recommendation method according to another embodiment of the present application, which is applicable to an electronic device, and the business object recommendation method may include:

step S310: at least one candidate service object under the target service is obtained, and each candidate service object corresponds to at least one target label.

Step S320: and acquiring at least one reference label generated in the reference service by the service user of the target service.

Step S330: and determining the label number of the target label in the at least one reference label from the at least one target label.

In some embodiments, after obtaining the at least one target tag and the at least one reference tag, the electronic device may also determine the recommendation score of each candidate business object according to the tag similarity between the at least one target tag and the at least one reference tag. Specifically, the number of tags of the target tag existing in the at least one reference tag may be determined from the at least one target tag, so as to determine the number of the same or similar interest/preference tags existing in different services of the reference service and the target service for the user, and thus the recommendation score of each candidate service object may be determined according to the number of the same or similar interest/preference.

Specifically, the electronic device may determine, for each candidate business object, at least one target tag corresponding to the candidate business object, and determine, for the at least one target tag, the number of tags of the target tags existing in the at least one reference tag, so as to obtain the number of tags corresponding to each candidate business object. For each target tag in the at least one target tag, whether the target tag exists in the at least one reference tag is determined, and if the target tag exists in the at least one reference tag, the number of the tags is increased by 1.

Step S340: and determining the recommendation score of each candidate business object according to the number of the labels.

In some embodiments, the recommendation score of each candidate business object may be determined according to the number of tags corresponding to each candidate business object. As one mode, the larger the number of tags corresponding to the candidate service object is, the larger the recommendation score of the candidate service object may be determined, and the smaller the number of tags corresponding to the candidate service object is, the smaller the recommendation score of the candidate service object may be determined.

It can be understood that, when the number of the target tags in the at least one target tag is greater, the target tag of the candidate service object is shown to be more likely to appear in the reference tags generated in the reference service by the user, that is, the more the target tag of the candidate service object is matched with the reference tags generated in the reference service by the user, the candidate service object determined by the target service can be considered to substantially reflect the actual interest preference of the user, and the recommendation of the candidate service object corresponding to the target tag is accurate, so that the recommendation priority of the candidate service object can be put forward to correspond to a higher recommendation priority. That is, when the recommendation score of the candidate service object is relatively large, the recommendation priority is higher.

Conversely, when the number of the target tags existing in the at least one reference tag is smaller from among the at least one target tag, it indicates that the smaller the probability that the target tag of the candidate service object appears in the reference tag generated in the reference service by the user is, that is, the more the target tag of the candidate service object is not matched with the reference tag generated in the reference service by the user, it may be considered that the candidate service object determined by the target service does not completely reflect the actual interest preference of the user, and the recommendation of the candidate service object corresponding to the target tag is not very accurate, so that the recommendation priority of the candidate service object may be set back to correspond to a lower recommendation priority, that is, the lower the recommendation score of the candidate service object is, the lower the recommendation priority is.

For example, in the above example, the number of matching tags of the candidate data recom _ book _ aa is relatively large (2), the score is high, and is assumed to be 10; the number of the match tags of the recom _ book _ bb is relatively small (0), the score is low, and the score is assumed to be 5; the higher the score, the greater the probability of acceptance by the user. So that the candidate book recom _ book _ aa can be subsequently preferentially selected for recommendation to the User 1.

Step S350: and according to the recommendation score, determining a target business object from the at least one candidate business object and recommending the target business object to the business user.

In the embodiment of the present application, step S350 may refer to the foregoing embodiments, and is not described herein again.

The service object recommendation method provided by the embodiment of the application obtains at least one candidate service object under a target service, wherein each candidate service object corresponds to at least one target tag, and obtains at least one reference tag generated by a service user of the target service in a reference service, wherein the reference service is different from the target service, the reference tag is determined according to the attention behavior of the service user in the reference service, then the number of tags of the target tag existing in the at least one reference tag is determined from the at least one target tag, the recommendation score of each candidate service object is determined according to the number of tags, and the target service object recommended to the service user is determined and recommended from the at least one candidate service object according to the recommendation score. According to the method and the device, the attention behaviors of the user under different service types are subjected to labeling processing, and the interest and the hobbies of the user under different service types can be obtained through classification, so that when the service object under the target service is recommended to the service user under the target service, the acceptance degree of the service user to each candidate service object can be comprehensively evaluated according to the similar interest and hobbies of the service user under other service types, and the accurate recommendation of the service object aiming at the service user can be realized according to the evaluation result.

Referring to fig. 8, fig. 8 is a schematic flowchart illustrating a business object recommendation method according to still another embodiment of the present application, which is applicable to an electronic device, and the business object recommendation method may include:

step S410: at least one candidate service object under the target service is obtained, and each candidate service object corresponds to at least one target label.

Step S420: and acquiring at least one reference label generated in the reference service by the service user of the target service.

Step S430: and acquiring the service similarity between the target service and the reference service.

It is understood that, when the service correlation degree of the reference service and the target service is larger, it is considered that the possibility that the users of the two services are interested in preferring similar items is larger. Therefore, in some embodiments, the recommendation scores of the candidate business objects may also be comprehensively evaluated in combination with the business relevance of the reference business and the target business.

Due to different services, user groups (user images) of the services have certain differences, and when the difference between the user groups of the two services is smaller, namely, the user groups of the two services are similar, the two services can be considered to be similar, namely, the correlation between the two services is larger. Therefore, in some embodiments, referring to fig. 9, step S430 may include:

step S431: and determining the target user attribute under the target service and the reference user attribute under the reference service.

The user attribute is information describing a corresponding user, and may include static attributes such as user age, user gender, user geographic location, user academic history, love and marriage status, user paying or not, used terminal equipment and the like, and may also include dynamic attributes such as user consumption level, access time, access duration and the like. The specific user attribute types are not limited herein.

In some embodiments, the user attributes that may be set correspondingly to different services may be the same, may also be different, or may also be partially the same, which is not limited herein. In the embodiment of the application, the electronic device may determine the target user attribute of the service user in the target service and the reference user attribute of the service user in the reference service, respectively. The target user attribute may be a user attribute set correspondingly under the target service, and the reference user attribute may be a user attribute set correspondingly under the reference service.

Step S432: and determining the service similarity of the target service and the reference service according to the matching degree of the target user attribute and the reference user attribute.

In some embodiments, after acquiring the target user attribute in the target service and the reference user attribute in the reference service, the electronic device may determine the service similarity between the target service and the reference service according to the matching degree of the target user attribute and the reference user attribute.

In some embodiments, the service similarity between the target service and the reference service may be determined according to the distribution of the target user attribute and each of the reference user attributes. Since there may be a case where the target user attribute is different from the reference user attribute, some target user attributes under the target service may not exist in the reference service, and the distribution condition cannot be determined for the non-existing user attributes, as an implementation, the service similarity between the target service and the reference service may be determined according to the distribution condition of the common attribute in the target user attribute and the reference user attribute. Specifically, referring to fig. 10, step S432 may include:

step S4321: and acquiring the common user attribute in the target user attribute and the reference user attribute.

As one mode, the electronic device may determine, for each target user attribute of the target service, whether the target user attribute exists in the reference user attribute of the reference service, and if the target user attribute exists, the target user attribute may be considered as a common user attribute in the target user attribute and the reference user attribute, so that all common user attributes in the target user attribute and the reference user attribute may be obtained. The number of the common user attributes may be one or more. As another mode, the electronic device may also determine, for each reference user attribute of the reference service, whether the reference user attribute exists in the target user attribute of the target service, and if the reference user attribute exists, the reference user attribute may be considered as a common user attribute in the target user attribute and the reference user attribute.

Step S4322: determining a third importance score of the common user attribute in the target business and a fourth importance score of the common user attribute in the reference business.

In some embodiments, after obtaining the common user attribute of the target user attribute and the reference user attribute, a third importance score of the common user attribute in the target service and a fourth importance score of the common user attribute in the reference service may be determined. Thereby obtaining the importance degree of the user attribute shared by the reference service and the target service in different services.

In a service, the importance score of a user attribute may be obtained by first obtaining the total number of service users in the service, then obtaining all service users with the user attribute, obtaining the ratio of all service users with the user attribute to the total number of service users, and then taking the ratio of each user attribute as the importance score of each user attribute in the service. That is, the importance score of a certain user attribute is equal to the number of service users/the total number of service users of a certain user attribute.

Specifically, for each common user attribute, a ratio of all service users having the common user attribute in the target service to a total number of service users in the target service may be obtained as a third importance score of the common user attribute in the target service, and a ratio of all service users having the common user attribute in the reference service to a total number of service users in the reference service may be obtained as a fourth importance score of the common user attribute in the reference service. Thereby, the importance scores of each common user attribute in the target service and the reference service can be obtained.

Step S4323: and determining the business similarity of the target business and the reference business according to the third importance score and the fourth importance score.

The service similarity between the target service and the reference service can be obtained according to a cosine similarity algorithm, and the cosine similarity algorithm can use a cosine value of an included angle between two vectors in a vector space as a measure for the difference between the two individuals. The cosine similarity algorithm is as follows:

wherein, the common user attribute shared by the service users in the target service X and the reference service Y is n, the value of i is the ith common user attribute, X_iA third importance score in target service X, Y, for the ith common user attribute_iA fourth importance score in reference service Y for the ith common user attribute. For example, the common user attribute is the duration of the subject science, the number of service users having the user attribute of the subject science in the target service X is 1500 (total number of service users 10000), and the number of service users in the service Y is 10000If the number of service users having the subject calendar user attribute is 2000 (total number of service users 20000), x (i) is 0.15, and y (i) is 0.10.

The service correlation degree cos theta can be calculated according to the formula, and the value range of the service correlation degree cos theta is 0-1. It is understood that the larger the value, the closer to 1, the closer to 0 degree the included angle is, and the more similar the target service X and the reference service Y are, the greater the value of mutual reference.

Step S440: and determining the recommendation score of each candidate business object according to the matching degree of the at least one reference label and the at least one target label and the business similarity.

In some embodiments, after determining the service similarity between the target service and the reference service, the recommendation score of each candidate service object may be determined according to the matching degree between the at least one reference tag and the at least one target tag and the service similarity. And when the business similarity is higher and the matching degree is higher, the recommendation score of the candidate business object corresponding to the at least one target label is higher.

In some embodiments, the matching degree of the at least one reference tag and the at least one target tag may be the matching degree of the first importance score and the second importance score of the previous embodiments. As one way, the service similarity may be subjected to a ratio operation with a difference between the first importance score and the second importance score to obtain a recommendation score of each candidate service object. The specific recommendation score calculation formula may be:

score＝r/∑(W(Xi)-W(Yi))²，i＝1...n

the target tags corresponding to the candidate service objects are n, the service correlation degree between the target service X and the reference service Y is r, the value of i is the ith target tag of the candidate service object, w (xi) is the second importance score of the ith target tag in the target service X, and w (yi) is the first importance score of the ith target tag in the reference service Y.

It is understood that, when the business similarity is higher and the difference between the first importance score and the second importance score is smaller, i.e. the numerator is larger and the denominator is smaller, the score of the overall obtained recommendation score is larger. That is to say, the more similar the target service X and the reference service Y are, the greater the possibility that the users of the two services are interested in preferring similar articles is, the higher the reference value of the reference service is, and meanwhile, the more the importance degrees of the target tags corresponding to the candidate service objects in different services are the same or similar, the greater the probability of being accepted by the users is, the higher the recommendation accuracy of the candidate service objects is, and therefore, the higher the recommendation score of the candidate service object is determined to be.

For example, in the foregoing example, assuming that the correlation between the reference service a and the target service B is 0.8, the electronic device may obtain the recommendation score of the candidate book recom _ book _ aa as follows: 0.8/((0.3-0.25)²+(0.1-0.15)²) 160, the recommendation score of the candidate book recom _ book _ bb is: 0.8/((0.1-0.01)²) 98.8. It can be seen that the recommendation score of the candidate book recom _ book _ aa is relatively large, and the probability that the candidate book recom _ book _ aa is accepted by the user is considered to be higher. So that the candidate book recom _ book _ aa can be subsequently preferentially selected for recommendation to the User 1.

Step S450: and according to the recommendation score, determining a target business object from the at least one candidate business object and recommending the target business object to the business user.

In the embodiment of the present application, step S450 can refer to the foregoing embodiments, and is not described herein again.

The service object recommendation method provided by the embodiment of the application obtains at least one candidate service object under a target service, wherein each candidate service object corresponds to at least one target tag, and obtains at least one reference tag generated by a service user of the target service in a reference service, wherein the reference service is different from the target service, the reference tag is determined according to the attention behavior of the service user in the reference service, then obtains the service similarity of the target service and the reference service, so as to determine the recommendation score of each candidate service object according to the matching degree of the at least one reference tag and the at least one target tag and the service similarity, and finally determines and recommends the target service object recommended to the service user from the at least one candidate service object according to the recommendation score. The recommendation score of the candidate business object corresponding to the at least one target label is higher when the business similarity is higher and the matching degree of the at least one reference label and the at least one target label is higher. According to the method and the device, the attention behaviors of the user under different service types are labeled, the interest and the hobbies of the user under different service types can be obtained through classification, so that when the service object under the target service is recommended to the service user under the target service, the acceptance degree of the service user to each candidate service object can be comprehensively evaluated according to the similar interest and hobbies of the service user under other service types and the service correlation degree of the target service and other service types, and the accurate recommendation of the service object aiming at the service user can be realized according to the evaluation result.

Referring to fig. 11, fig. 11 is a schematic flowchart illustrating a flow of a service object recommendation method provided in an embodiment of the present application, where the method is applicable to an electronic device, and the service object recommendation method may include:

step S510: and determining similar service users of each service user according to the user attribute of each service user under the target service.

Since users with similar user attributes (similar user figures) are more likely to like similar items, in some embodiments, the electronic device may recommend the business object to the user based on the user attribute similarity. Specifically, the electronic device may determine similar service users of each service user under the target service according to the user attribute of each service user under the target service, so as to recommend the service users under the target service according to the service objects, which are interested and preferred by the similar service users.

In some embodiments, the user attribute of each service user in the target service may be obtained, the attribute similarity between all service users in the target service may be determined, and a similarity matrix may be obtained, which may be updated at regular time. When the target service needs to recommend content to the designated service user under the target service, a similar service user list of the designated service user can be obtained according to the similarity matrix, so as to determine the similar service user of the designated service user according to the similar service user category.

As a mode, the attribute similarity between two users may be obtained according to a cosine similarity algorithm, which may use a cosine value of an included angle between two vectors in a vector space as a measure of the difference between two individuals. Specifically, for users a and b, the cosine similarity algorithm is as follows:

wherein, the number of user attributes under the target service is n; the value of i is the ith user attribute under the target service; x_iIs used for indicating whether the ith user attribute exists in the user a or not, and X is the case_iHas a value of 1, in the absence of X_iIs 0; y is_iIs used for indicating whether the ith user attribute exists in the user b or not, and Y is used if the ith user attribute exists in the user b_iHas a value of 1, in the absence Y_iThe value of (d) is 0.

The attribute similarity cos theta between the two users can be calculated according to the formula, and the value range of the attribute similarity cos theta is 0-1. It will be appreciated that a larger value, closer to 1, indicates a closer angle to 0 degrees, and users a and b may be considered to be more similar and worth referencing each other.

In some embodiments, the service users with the attribute similarity value smaller than the preset similarity value may all be used as similar service users of the designated service user, or the similar service users may be ranked according to the sequence of the attribute similarity values from small to large in the list of similar service users, and then the preset number of service users with higher similarity ranked in the front may be obtained as similar service users of the designated service user. The preset similarity value and the preset number can be set according to actual requirements, and are not limited herein. For example, the preset similarity value may be 0.3 and the preset number may be 100.

In some embodiments, the service object to be recommended may be determined from the service objects concerned by the similar service users in the target service according to the number of the service objects carried in the request parameter. The request parameter may be a request initiated by a target service accessing to the general attention system of the application, and may include: the recommended specified service user ID, the similarity ranking threshold (namely the preset number), and the recommended service object number.

Step S520: and determining at least one candidate business object under the target business according to the business object concerned by the similar business user in the target business, wherein each candidate business object corresponds to at least one target label.

In some embodiments, after determining the similar service user of each service user, a service object concerned by the similar service user in the target service may be obtained as at least one candidate service object under the target service, so as to perform further recommendation score determination in the foregoing embodiments.

It can be understood that, since the business object recommended to a specific business user is determined only by relying on the similarity of attributes between all business users under the target business, it is likely not to completely reflect the actual interest preference of the specific business user. Therefore, in the embodiment of the application, the interest preference of the specified service user reflected in other services can be obtained, so as to synthesize the interest preferences of the user among different services to accurately recommend the service object under the target service to the specified service user

Step S530: and acquiring at least one reference label generated in a reference service by a service user of the target service, wherein the reference service is different from the target service, and the reference label is determined according to the concerned behavior of the service user in the reference service.

Step S540: and determining a recommendation score of each candidate business object according to the at least one reference label and the at least one target label.

Step S550: and according to the recommendation score, determining a target business object from the at least one candidate business object and recommending the target business object to the business user.

In the embodiment of the present application, steps S530 to S550 may refer to the foregoing embodiments, and are not described herein again.

The service object recommendation method provided by the embodiment of the application determines similar service users of each service user through the user attribute of each service user under the target service, so as to determine at least one candidate service object under the target service according to the service object concerned by the similar service user in the target service, wherein each candidate service object corresponds to at least one target label, and obtains at least one reference label generated in the reference service by the service user of the target service, wherein the reference service is different from the target service, the reference label is determined according to the concerned behavior of the service user in the reference service, then determining a recommendation score of each candidate business object according to the at least one reference label and the at least one target label, and determining and recommending a target service object recommended to the service user from the at least one candidate service object according to the recommendation score. According to the method and the device, the attention behaviors of the users under different service types are subjected to labeling processing, and the interest of the users under different service types can be obtained in a classifying manner, so that when the candidate service object is recommended to the service user according to the service object which is concerned by the similar service user of each service user in the target service, the acceptance degree of the service user on the candidate service object can be further comprehensively evaluated according to the interest of the service user under other service types, and the accurate recommendation of the service object aiming at the service user can be realized according to the evaluation result.

Referring to fig. 12, fig. 12 is a block diagram illustrating a structure of a service object recommendation apparatus 400 according to an embodiment of the present application, where the service object recommendation apparatus 400 is applied to an electronic device. The business object recommendation apparatus 400 includes: a candidate object acquisition module 410, a reference tag acquisition module 420, a recommendation score evaluation module 430, and a target object recommendation module 440. The candidate object obtaining module 410 is configured to obtain at least one candidate service object under a target service, where each candidate service object corresponds to at least one target tag; the reference tag obtaining module 420 is configured to obtain at least one reference tag generated in a reference service by a service user of the target service, where the reference service is different from the target service, and the reference tag is determined according to a concerned behavior of the service user in the reference service; the recommendation score evaluation module 430 is configured to determine a recommendation score for each of the candidate business objects according to the at least one reference tag and the at least one target tag; the target object recommending module 440 is configured to determine a target service object from the at least one candidate service object according to the recommendation score and recommend the target service object to the service user.

In some embodiments, the recommendation score evaluation module 430 may include: a tag matching unit configured to determine, for each of the target tags, a specified reference tag that matches the target tag from the at least one reference tag; a first score obtaining unit, configured to obtain an importance score of the specified reference tag in the reference service as a first importance score of the target tag in the reference service; the second score acquisition unit is used for determining a second importance score of the target label in the target business; a score determining unit, configured to determine a recommendation score for each candidate business object according to the first importance score and the second importance score.

In some embodiments, the business object recommending apparatus 400 may further include: and a default score obtaining module, configured to, for each target tag, obtain a preset default importance score as a first importance score of the target tag in the reference service when a specified reference tag matching the target tag does not exist in the at least one reference tag.

In other embodiments, the recommendation score evaluation module 430 may also be specifically configured to: determining the label number of a target label in the at least one reference label from the at least one target label; and determining the recommendation score of each candidate business object according to the number of the labels.

In still other embodiments, the recommendation score evaluation module 430 may also include: the service evaluation unit is used for acquiring the service similarity between the target service and the reference service; and the comprehensive evaluation unit is used for determining the recommendation score of each candidate business object according to the matching degree of the at least one reference label and the at least one target label and the business similarity, wherein the recommendation score of the candidate business object corresponding to the at least one target label is higher when the business similarity is higher and the matching degree is higher.

In some embodiments, the service evaluation unit may include: an attribute determining subunit, configured to determine a target user attribute under the target service and a reference user attribute under the reference service; and the attribute matching subunit is used for determining the service similarity between the target service and the reference service according to the matching degree of the target user attribute and the reference user attribute.

In some embodiments, the attribute matching subunit may be specifically configured to: acquiring a common user attribute in the target user attribute and the reference user attribute; determining a third importance score for the common user attribute in the target business and a fourth importance score for the common user attribute in the reference business; and determining the business similarity of the target business and the reference business according to the third importance score and the fourth importance score.

In some embodiments, the candidate acquisition module 410 may be specifically configured to: determining similar service users of each service user according to the user attribute of each service user under the target service; and determining at least one candidate service object under the target service according to the concerned service object of the similar service user in the target service.

The service object recommendation device provided in the embodiment of the present application is used to implement the corresponding service object recommendation method in the foregoing method embodiments, and has the beneficial effects of the corresponding method embodiments, which are not described herein again.

It can be clearly understood by those skilled in the art that, for convenience and brevity of description, the specific working processes of the above-described apparatuses and modules may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again.

In the several embodiments provided in the present application, the coupling or direct coupling or communication connection between the modules shown or discussed may be through some interfaces, and the indirect coupling or communication connection between the devices or modules may be in an electrical, mechanical or other form.

In addition, functional modules in the embodiments of the present application may be integrated into one processing module, or each of the modules may exist alone physically, or two or more modules are integrated into one module. The integrated module can be realized in a hardware mode, and can also be realized in a software functional module mode.

Referring to fig. 13, fig. 13 is a block diagram illustrating a structure of an electronic device according to an embodiment of the present disclosure. The electronic device 700 may be the terminal device, wherein the terminal device may be a user device capable of running an application, such as a smart phone, a tablet computer, a notebook computer, a desktop computer, and a wearable terminal device. The electronic device 700 may also be the server described above. The electronic device 700 in the present application may include one or more of the following components: a processor 710, a memory 720, and one or more applications, wherein the one or more applications may be stored in the memory 720 and configured to be executed by the one or more processors 710, the one or more programs configured to perform a method as described in the aforementioned method embodiments.

Processor 710 may include one or more processing cores. The processor 710 interfaces with various components throughout the electronic device 700 using various interfaces and circuitry to perform various functions of the electronic device 700 and process data by executing or executing instructions, programs, code sets, or instruction sets stored in the memory 720 and invoking data stored in the memory 720. Alternatively, the processor 710 may be implemented in hardware using at least one of Digital Signal Processing (DSP), Field-Programmable Gate Array (FPGA), and Programmable Logic Array (PLA). The processor 710 may integrate one or more of a Central Processing Unit (CPU), a Graphics Processing Unit (GPU), a modem, and the like. Wherein, the CPU mainly processes an operating system, a user interface, an application program and the like; the GPU is used for rendering and drawing display content; the modem is used to handle wireless communications. It is understood that the modem may not be integrated into the processor 710, but may be implemented by a communication chip.

The Memory 720 may include a Random Access Memory (RAM) or a Read-Only Memory (Read-Only Memory). The memory 720 may be used to store instructions, programs, code sets, or instruction sets. The memory 720 may include a stored program area and a stored data area, wherein the stored program area may store instructions for implementing an operating system, instructions for implementing at least one function (such as a touch function, a sound playing function, an image playing function, etc.), instructions for implementing various method embodiments described below, and the like. The storage data area may also store data created during use by the electronic device 700, and the like.

Those skilled in the art will appreciate that the structure shown in fig. 13 is a block diagram of only a portion of the structure relevant to the present application, and does not constitute a limitation on the electronic device to which the present application is applied, and a particular electronic device may include more or less components than those shown in the drawings, or combine certain components, or have a different arrangement of components.

Referring to fig. 14, a block diagram of a computer-readable storage medium according to an embodiment of the present disclosure is shown. The computer-readable storage medium 800 stores program code that can be called by a processor to execute the methods described in the above-described method embodiments.

The computer-readable storage medium 800 may be an electronic memory such as a flash memory, an EEPROM (electrically erasable and programmable read only memory), an EPROM, a hard disk, or a ROM. Alternatively, the computer-readable storage medium 800 includes a non-transitory computer-readable storage medium. The computer readable storage medium 800 has storage space for program code 810 for performing any of the method steps described above. The program code can be read from or written to one or more computer program products. The program code 810 may be compressed, for example, in a suitable form.

In the description herein, reference to the description of the term "one embodiment," "some embodiments," "an example," "a specific example," or "some examples," etc., means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the application. In this specification, the schematic representations of the terms used above are not necessarily intended to refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. Furthermore, various embodiments or examples and features of different embodiments or examples described in this specification can be combined and combined by one skilled in the art without contradiction.

Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit the same; although the present application has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; such modifications and substitutions do not necessarily depart from the spirit and scope of the corresponding technical solutions in the embodiments of the present application.

Claims

1. A business object recommendation method, characterized in that the method comprises:

acquiring at least one candidate service object under a target service, wherein each candidate service object corresponds to at least one target label;

acquiring at least one reference label generated in a reference service by a service user of the target service, wherein the reference service is different from the target service, and the reference label is determined according to the concerned behavior of the service user in the reference service;

determining a recommendation score for each of the candidate business objects based on the at least one reference tag and the at least one target tag;

and according to the recommendation score, determining a target business object from the at least one candidate business object and recommending the target business object to the business user.

2. The method of claim 1, wherein determining a recommendation score for each of the candidate business objects based on the at least one reference tag and the at least one target tag comprises:

for each of the target tags, determining a designated reference tag matching the target tag from the at least one reference tag;

acquiring an importance score of the specified reference label in the reference service as a first importance score of the target label in the reference service;

determining a second importance score of the target tag in the target business;

determining a recommendation score for each of the candidate business objects based on the first importance score and the second importance score.

3. The method of claim 2, wherein prior to said determining a recommendation score for each of said candidate business objects based on said first importance score and said second importance score, the method further comprises:

for each target label, when a specified reference label matched with the target label does not exist in the at least one reference label, acquiring a preset default importance score as a first importance score of the target label in the reference service.

4. The method of claim 1, wherein determining a recommendation score for each of the candidate business objects based on the at least one reference tag and the at least one target tag comprises:

determining the label number of a target label in the at least one reference label from the at least one target label;

and determining the recommendation score of each candidate business object according to the number of the labels.

5. The method of claim 1, wherein determining a recommendation score for each of the candidate business objects based on the at least one reference tag and the at least one target tag comprises:

acquiring the service similarity between the target service and the reference service;

and determining the recommendation score of each candidate business object according to the matching degree of the at least one reference label and the at least one target label and the business similarity, wherein the higher the business similarity is and the higher the matching degree is, the higher the recommendation score of the candidate business object corresponding to the at least one target label is.

6. The method of claim 5, wherein the obtaining the service similarity between the target service and the reference service comprises:

determining the target user attribute under the target service and the reference user attribute under the reference service;

and determining the service similarity of the target service and the reference service according to the matching degree of the target user attribute and the reference user attribute.

7. The method of claim 6, wherein the determining the service similarity between the target service and the reference service according to the matching degree of the target user attribute and the reference user attribute comprises:

acquiring a common user attribute in the target user attribute and the reference user attribute;

determining a third importance score for the common user attribute in the target business and a fourth importance score for the common user attribute in the reference business;

and determining the business similarity of the target business and the reference business according to the third importance score and the fourth importance score.

8. The method according to any one of claims 1 to 7, wherein the obtaining at least one candidate service object under the target service comprises:

determining similar service users of each service user according to the user attribute of each service user under the target service;

and determining at least one candidate service object under the target service according to the concerned service object of the similar service user in the target service.

9. A business object recommendation apparatus, the apparatus comprising:

the candidate object acquisition module is used for acquiring at least one candidate service object under a target service, and each candidate service object corresponds to at least one target label;

a reference tag obtaining module, configured to obtain at least one reference tag generated in a reference service by a service user of the target service, where the reference service is different from the target service, and the reference tag is determined according to a behavior of the service user concerning the reference service;

a recommendation score evaluation module for determining a recommendation score for each of the candidate business objects according to the at least one reference tag and the at least one target tag;

and the target object recommending module is used for determining a target service object from the at least one candidate service object according to the recommending score and recommending the target service object to the service user.

10. An electronic device, comprising:

one or more processors;

a memory;

one or more applications, wherein the one or more applications are stored in the memory and configured to be executed by the one or more processors, the one or more applications configured to perform the method of any of claims 1-8.

11. A computer-readable storage medium, having stored thereon program code that can be invoked by a processor to perform the method according to any one of claims 1 to 8.