Business recommendation method, device, equipment and storage medium

Abstract

The present application discloses a service recommendation method, apparatus, device and storage medium, which relates to the field of data processing technology and is used to improve the accuracy and efficiency of recommending services to users. The method includes: based on the target data information of each user among multiple users, clustering multiple users through a preset algorithm and multiple preset rules to obtain at least one feature tag corresponding to each user; based on the target data information of each user among multiple users, determining the similarity between the target data information of any two users through a preset knowledge graph model; based on the similarity between the target data information of any two users, determining at least one weight value corresponding to each feature tag of at least one feature tag corresponding to each user; and according to the weight value corresponding to each feature tag of at least one feature tag corresponding to any user among multiple users, determining the target service with the greatest demand from a plurality of services.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a service recommendation method, apparatus, device, and storage medium. Background Art

[0002] As operators' businesses develop, user demand for package services continues to increase and diversify. Therefore, it's necessary to recommend appropriate packages for different users. Currently, when recommending packages to users, collaborative filtering algorithms can be used to predict historical data on users' use of different package products (i.e., package services). This can then be used to obtain user ratings for different package products and recommend the highest-rated packages to users.

[0003] In the above method, when a user has a limited number of package products, data sparsity can lead to low accuracy in recommended package products. Furthermore, when recommending package services to new users, the accuracy of recommended packages is also low due to the limited user data collected. Consequently, service recommendations to users are less accurate and inefficient. Summary of the Invention

[0004] The present application provides a service recommendation method, apparatus, device and storage medium for solving the problem of low accuracy in recommending services to users when the user has used fewer types of package services in the past, thereby improving the accuracy and efficiency of recommending services to users.

[0005] To achieve the above objectives, this application adopts the following technical solutions:

[0006] In a first aspect, a service recommendation method is provided, the method comprising: based on target data information of each user among a plurality of users, clustering the plurality of users through a preset algorithm and a plurality of preset rules to obtain at least one feature tag corresponding to each of the plurality of users, the target data information comprising at least one of the following: basic information, service usage information, basic information of activated services, one preset rule corresponding to one feature tag, one feature tag corresponding to at least one user, and the feature tags comprising at least one of the following: a service value tag, a service function tag, and a user status tag; based on the target data information of each user among the plurality of users, determining the similarity between the target data information of any two users through a preset knowledge graph model, the knowledge graph model being used to associate the plurality of target data information; based on the similarity between the target data information of any two users, determining at least one weight value corresponding to each feature tag of at least one feature tag corresponding to each user among the plurality of users, the weight value being used to indicate the degree of demand of the user corresponding to the feature tag for one of the plurality of preset services; and determining, from the plurality of services, the target service for which any user has the greatest demand, according to the weight value corresponding to each feature tag of at least one feature tag corresponding to any user among the plurality of users.

[0007] In one possible implementation, the method further includes: obtaining original data information of each user among multiple users; performing data preprocessing on the original data information of each user among the multiple users to obtain target data information of each user, where the data preprocessing includes: data cleaning, data error correction, and data unification.

[0008] In one possible implementation, based on the target data information of each user among multiple users, the similarity between the target data information of any two users is determined through a preset knowledge graph model, including: based on the target data information of each user among multiple users, a target vector corresponding to the target data information of any two users is constructed through a knowledge graph model; the target vector corresponding to the target data information of any two users is introduced into a target hyperplane, and the target data information of each user is represented in multiple relationship planes in the target hyperplane, where the target hyperplane is constructed based on multiple users, multiple services, and user status labels, and the user status labels include at least one of the following: users to be lost, stable users, and users with service changes; based on the target hyperplane and the optimal transmission algorithm, the similarity between the target data information of any two users among the multiple users is determined.

[0009] In one possible implementation, based on the similarity between the target data information of any two users, at least one weight value corresponding to each feature tag of at least one feature tag corresponding to each user of a plurality of users is determined, including: based on the similarity between the target data information of any two users of the plurality of users, determining the association between any feature tag of the plurality of feature tags corresponding to one of the two users and a first service activated by the other user; and determining the weight value corresponding to any feature tag and the first service according to the association between any feature tag and the first service.

[0010] In a second aspect, a service recommendation device is provided, which includes: a processing unit and a determination unit; the processing unit is used to cluster the multiple users based on target data information of each user in the multiple users through a preset algorithm and multiple preset rules to obtain at least one feature label corresponding to each user in the multiple users, the target data information includes at least one of the following: basic information, service usage information, basic information of activated services, one preset rule corresponds to one feature label, one feature label corresponds to at least one user, and the feature label includes at least one of the following: service value label, service function label, user status label; the determination unit is used to cluster the multiple users based on the target data information of each user in the multiple users Data information, determines the similarity between the target data information of any two users through a preset knowledge graph model, and the knowledge graph model is used to associate multiple target data information; the determination unit is also used to determine, based on the similarity between the target data information of any two users, at least one weight value corresponding to each feature tag in at least one feature tag corresponding to each user of the multiple users, and the weight value is used to indicate the demand of the user corresponding to the feature tag for one of the preset multiple services; the determination unit is also used to determine the target service with the greatest demand from any user from the multiple services based on the weight value corresponding to each feature tag in at least one feature tag corresponding to any user of the multiple users.

[0011] In one possible implementation, the service recommendation device further includes an acquisition unit; the acquisition unit is used to obtain original data information of each user among a plurality of users; the processing unit is further used to perform data preprocessing on the original data information of each user among the plurality of users to obtain target data information of each user, where the data preprocessing includes: data cleaning, data error correction, and data unification.

[0012] In one possible implementation, the processing unit is further used to construct a target vector corresponding to the target data information of any two users through a knowledge graph model based on the target data information of each user among multiple users; the processing unit is further used to introduce the target vector corresponding to the target data information of any two users into a target hyperplane, and represent the target data information of each user in multiple relationship planes in the target hyperplane, where the target hyperplane is constructed based on multiple users, multiple services, and user status labels, and the user status labels include at least one of the following: users to be lost, stable users, and users with service changes; the determination unit is further used to determine the similarity between the target data information of any two users among multiple users based on the target hyperplane and the optimal transmission algorithm.

[0013] In one possible implementation, the determination unit is further used to determine, based on the similarity between the target data information of any two users among the multiple users, the association between any one of the multiple feature tags corresponding to one of the two users and the first service activated by the other user; the determination unit is further used to determine the weight value corresponding to any feature tag and the first service based on the association between any one of the feature tags and the first service.

[0014] In a third aspect, an electronic device comprises: a processor and a memory; wherein the memory is used to store one or more programs, and the one or more programs include computer execution instructions. When the electronic device is running, the processor executes the computer execution instructions stored in the memory to enable the electronic device to perform a business recommendation method as in the first aspect.

[0015] In a fourth aspect, a computer-readable storage medium storing one or more programs is provided. The one or more programs include instructions. When executed by a computer, the instructions enable the computer to execute a business recommendation method as described in the first aspect.

[0016] The present application provides a service recommendation method, apparatus, device, and storage medium for use in scenarios involving recommending services to users. When recommending services to users, the method can cluster multiple users based on target data corresponding to each user using a preset algorithm and multiple preset rules to obtain a feature label corresponding to each user. Furthermore, the multiple target data corresponding to the multiple users can be associated using a preset knowledge graph model to determine the similarity between the target data of any two users. The degree of similarity between any two data pieces indicates the closeness of the association between the two data pieces. Furthermore, based on the similarity between the target data of any two users, a weight value for the feature label corresponding to each user is further determined to identify the target service with the highest user demand from multiple services based on the weight value. Specifically, the greater the similarity between the two data pieces, the higher the weight value for the feature label, the greater the user's demand for the service, and the more appropriate the service recommended to the user. This solves the problem of low accuracy in service recommendations when a user has a limited number of services historically used, thereby improving the accuracy and efficiency of service recommendations. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 A schematic diagram of the structure of a business recommendation system provided in an embodiment of the present application;

[0018] Figure 2 A process diagram of a business recommendation method provided in an embodiment of this application Figure 1 ;

[0019] Figure 3 A schematic diagram of a process for recommending a communication user service according to an embodiment of the present application;

[0020] Figure 4 A process diagram of a business recommendation method provided in an embodiment of this application Figure 2 ;

[0021] Figure 5 A process diagram of a business recommendation method provided in an embodiment of this application Figure 3 ;

[0022] Figure 6 A process diagram of a business recommendation method provided in an embodiment of this application Figure 4 ;

[0023] Figure 7 A schematic diagram of the structure of a business recommendation device provided in an embodiment of the present application;

[0024] Figure 8 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0025] The technical solutions in the embodiments of the present application will be described below in conjunction with the drawings in the embodiments of the present application.

[0026] In the description of this application, unless otherwise specified, " / " means "or", for example, A / B can mean A or B. "And / or" in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone. In addition, "at least one" and "a plurality of" refer to two or more. Words such as "first" and "second" do not limit the quantity and execution order, and words such as "first" and "second" do not limit them to be different.

[0027] As operators' businesses develop, the market is approaching saturation, with fewer new users. However, user demand for package services is increasing and diversifying. To retain existing users, it's necessary to launch customized packages tailored to each user. Currently, when recommending package services to users, collaborative filtering algorithms can be used to predict historical user data on different package products (i.e., package services). This can then be used to determine user ratings for each package, and recommend the highest-rated package.

[0028] In the above method, when a user uses a limited number of packages, data sparsity can lead to low accuracy in recommended packages. Furthermore, when recommending packages to new users, the cold start problem, caused by limited user data, can also reduce the accuracy of recommended packages. Furthermore, the diversity and interpretability of the recommendation results cannot be guaranteed.

[0029] This application provides a service recommendation method. When recommending services to users, the method can cluster multiple users based on target data corresponding to each user using a preset algorithm and multiple preset rules to obtain feature labels corresponding to each user. Furthermore, the method can associate multiple target data corresponding to the multiple users using a preset knowledge graph model to determine the similarity between the target data of any two users. The similarity between any two data pieces indicates the closeness of the association between the two data pieces. Furthermore, based on the similarity between the target data of any two users, a weight value for the feature label corresponding to each user is further determined to identify the target service with the highest user demand from multiple services based on the weight value. Specifically, the greater the similarity between the two data pieces, the higher the weight value for the feature label, the greater the user's demand for the service, and the more appropriate the service recommended to the user. This method solves the problem of low accuracy in service recommendations when a user has a limited number of services historically used, thereby improving the accuracy and efficiency of service recommendations.

[0030] The embodiment of the present application provides a service recommendation method, which can be applied to a service recommendation system. Figure 1 Figure 2 shows the structural diagram of the business recommendation system. Figure 1 As shown, the service recommendation system 10 includes: an electronic device 11 and an operator server 12 .

[0031] Among them, the electronic device 11 is used to obtain target data information of each user among multiple users from the operator server 12, cluster multiple users based on the target data information of multiple users to obtain feature labels of each user, determine the similarity between the target data information of any two users based on the target data information of multiple users, determine the weight value corresponding to the feature label based on the similarity between the target data information of any two users, and determine the target service with the highest user demand from multiple services according to the weight value corresponding to the feature label.

[0032] The operator server 12 is used to send target data information of each user among multiple users to the electronic device 11, so that the electronic device 11 can determine the target service with the greatest user demand and recommend it to the user.

[0033] The following describes a service recommendation method provided by an embodiment of the present application in conjunction with the accompanying drawings. Figure 2 As shown, an embodiment of the present application provides a service recommendation method, which is applied to an electronic device. The method includes S201-S204:

[0034] S201 : Based on target data information of each user among a plurality of users, clustering processing is performed on the plurality of users by a preset algorithm and a plurality of preset rules to obtain at least one feature label corresponding to each user among the plurality of users.

[0035] Among them, the target data information includes at least one of the following: basic information, business usage information, basic information of activated businesses, one preset rule corresponds to one feature tag, one feature tag corresponds to at least one user, and the feature tag includes at least one of the following: business value tag, business function tag, user status tag.

[0036] It can be understood that the electronic device can cluster multiple users based on at least one target data information of each user among multiple users, including basic information, business usage information, and basic information of activated services, through a preset algorithm and multiple preset rules, to obtain at least one feature label corresponding to each user among the multiple users, including at least one of a business value label, a business function label, and a user status label.

[0037] Optionally, the basic information may include the user's gender, age, and place of origin; the service usage information may include the user's average monthly number of calls, average monthly call duration, average monthly number of text messages, average monthly Internet traffic, the name of the currently used package, the number of basic services activated, and the number of value-added services activated in the past year; the basic information of the activated services may include the price of the currently used package, the basic functions of the currently used package (including fixed call duration, fixed number of text messages, fixed traffic, variable traffic), and the value-added functions of the currently used package (i.e., specific functions, such as positioning function, student worry-free function, fifth-generation mobile communication technology (5G) wireless dedicated line function, and video dedicated line function).

[0038] Optionally, based on each user's basic information and service usage information, a preset algorithm and multiple preset rules may be used to cluster multiple users to obtain at least one user status label corresponding to each user. Based on each user's basic information, the price of the currently used package, and the basic functions of the currently used package, a preset algorithm and multiple preset rules may be used to cluster multiple users to obtain at least one service value label corresponding to each user. Based on each user's basic information and the value-added functions of the currently used package, a preset algorithm and multiple preset rules may be used to cluster multiple users to obtain at least one service function label corresponding to each user.

[0039] Optionally, at least one user status tag may include users about to churn, stable users, and users with changing needs. At least one service function tag may include potential users. Multiple preset rules may include Rule 1, Rule 2, Rule 3, and Rule 4. Rule 1 may be for users whose average monthly Internet traffic is less than a preset usage and whose average monthly call duration is less than a preset duration; Rule 2 may be for users whose average monthly Internet traffic is greater than a preset usage and whose average monthly call duration is greater than a preset duration; Rule 3 may be for users whose average monthly Internet traffic is greater than a preset usage and whose average monthly call duration is less than a preset duration, or users whose average monthly Internet traffic is less than a preset usage and whose average monthly call duration is greater than a preset duration. Rule 4 may be for users who are older than a first preset age or younger than the first preset age and who need value-added services, or for users who are older than a third preset age and younger than a fourth preset age and who need value-added services.

[0040] Optionally, a data dashboard display of at least one feature tag corresponding to each user can be implemented through Python code.

[0041] Exemplarily, the preset algorithms may be classification algorithms and clustering algorithms. Users with an average monthly internet traffic of less than 1Gb and an average monthly call duration of less than 10 minutes may be considered potential users; users with an average monthly internet traffic of more than 1Gb and an average monthly call duration of more than 10 minutes may be considered stable users; users with an average monthly internet traffic of more than 1Gb and an average monthly call duration of less than 10 minutes, or users with an average monthly internet traffic of less than 1Gb and an average monthly call duration of more than 10 minutes may be considered users with changing needs; users aged 60 or older but younger than 6 who require positioning functions, or users aged 6 or older but younger than 18 who require the Student Worry-Free function may be considered potential users.

[0042] S202. Based on the target data information of each user among the multiple users, determine the similarity between the target data information of any two users through a preset knowledge graph model.

[0043] Among them, the knowledge graph model is used to associate multiple target data information.

[0044] It can be understood that the electronic device can determine the similarity between the target data information of any two users based on the target data information of each user among multiple users through a preset knowledge graph model for associating multiple target data information.

[0045] Optionally, knowledge extraction and knowledge fusion can be performed on the target data information of each user among multiple users through a preset knowledge graph and a generative adversarial network (GAN) to determine the similarity between the target data information of any two users.

[0046] S203: Determine at least one weight value corresponding to each feature tag in at least one feature tag corresponding to each user among the multiple users based on the similarity between the target data information of any two users.

[0047] The weight value is used to indicate the degree of demand of the user corresponding to the feature tag for one of the preset multiple services.

[0048] It can be understood that the electronic device can determine the weight value of the user's demand for one of the preset multiple services corresponding to at least one indication feature tag corresponding to each feature tag of at least one feature tag corresponding to each user among multiple users based on the similarity between the target data information of any two users.

[0049] Optionally, based on the similarity between the target data information of any two users, each feature tag in at least one feature tag corresponding to each user in the multiple users is differentially weighted through a collaborative filtering algorithm to obtain at least one weight value corresponding to each feature tag in at least one feature tag corresponding to each user in the multiple users.

[0050] S204: Determine, from the multiple services, a target service that is most in demand by any user according to a weight value corresponding to each feature tag in at least one feature tag corresponding to any user among the multiple users.

[0051] It can be understood that the electronic device can determine the target service most in demand by any user from among multiple services according to the weight value corresponding to each feature tag in at least one feature tag corresponding to any user among multiple users.

[0052] Optionally, a logistic regression prediction model can be used to determine the target service with the highest demand from multiple services based on the weight value corresponding to each of the at least one feature tag corresponding to each user. The service recommendation results are evaluated using the root mean square error parameter and the accuracy parameter of the recommendation results of the Top-N package recommendation evaluation index.

[0053] In one possible implementation, Figure 3As shown, a flow chart of a communication user service recommendation process is shown. The basic information, service usage information, and basic information of activated services of multiple users can be clustered to obtain various feature labels, and the basic information, service usage information, and basic information of activated services of multiple users can be subjected to knowledge extraction and knowledge fusion to obtain the similarity between the data information. Furthermore, based on the similarity between the data information, the weight values ​​of various feature labels compared to each of the multiple services are obtained, and the weight values ​​of various feature labels compared to each of the multiple services are sent to the prediction recommendation system. Through the prediction recommendation system, based on the basic information, service usage information, and basic information of activated services of any user, the feature label corresponding to the user is obtained. The target service corresponding to the user is determined based on the feature label corresponding to the user and the weight value of the feature label corresponding to the user compared to each of the multiple services.

[0054] In one possible implementation, multiple target data can be associated using a preset knowledge graph model to determine the similarity between any two users' target data. The higher the similarity between any two data, the closer the association. By integrating a generative adversarial network into the knowledge graph, deeper associations between multiple target data can be mined to compensate for the fragmentation caused by random sampling, improve the performance of the knowledge graph, and capture deeper user preferences through associations between multiple users' data. Weights corresponding to feature labels can be determined based on the similarity between any two users' target data. The greater the similarity, the higher the weight. The weights corresponding to the feature labels of any user can then be used to identify the target service with the highest demand from multiple services. The higher the weight, the greater the user's demand for that service. This allows for the most appropriate service to be recommended to the user, achieving optimal matching between the user and the service. This addresses the issue of low service recommendation accuracy when a user has a limited history of using a limited number of services, improving both the accuracy and efficiency of service recommendations.

[0055] In one design, Figure 4 As shown, the embodiment of the present application provides a service recommendation method, which further includes S301-S302:

[0056] S301: Obtain original data information of each user among multiple users.

[0057] S302: Perform data preprocessing on original data information of each user among the multiple users to obtain target data information of each user.

[0058] Among them, data preprocessing includes: data cleaning, data error correction, and data unification.

[0059] It can be understood that the original data information of each user among multiple users can be obtained, and the original data information of each user among the multiple users can be preprocessed including data cleaning, data error correction, and data unification to obtain the target data information of each user.

[0060] Optionally, data cleaning can include deleting data from the raw data that is irrelevant to basic information, service usage information, and basic information about activated services, as well as deleting duplicate data from the raw data. Data error correction can include investigating and correcting abnormal data in the raw data. Data unification can include integrating the raw data according to unified data standards and converting it into data suitable for sorting and storage. The target data information for each user obtained through data preprocessing can be imported and stored in a database.

[0061] In a possible implementation, target data information is obtained by performing data preprocessing on the original data information, which can provide a good data basis for subsequent clustering of multiple users to obtain at least one feature label corresponding to each of the multiple users.

[0062] In one design, Figure 5 As shown, the service recommendation method provided by the embodiment of the present application, the method in the above step S202 specifically includes S401-S403:

[0063] S401. Based on the target data information of each user among multiple users, a target vector corresponding to the target data information of any two users is constructed through a knowledge graph model.

[0064] It can be understood that based on the target data information of each user among multiple users, a target vector corresponding to the target data information of any two users can be constructed through a knowledge graph model.

[0065] Optionally, the basic information, business usage information, and basic information of activated businesses of any two users can be used as entity objects through the knowledge graph model to construct the target vector corresponding to the basic information and business usage information of any two users, the target vector corresponding to the basic information and the basic information of activated businesses, and the target vector corresponding to the business usage information and the basic information of activated businesses.

[0066] S402: introducing the target vector corresponding to the target data information of any two users into the target hyperplane, and representing the target data information of each user in multiple relationship planes in the target hyperplane.

[0067] The target hyperplane is constructed based on multiple users, multiple services, and user status labels. The user status labels include at least one of the following: users to be lost, stable users, and users with service changes.

[0068] It can be understood that the target vector corresponding to the target data information of any two users can be introduced into the target hyperplane constructed based on multiple users, multiple services, and user status labels, and the target data information of each user can be represented in multiple relationship planes in the target hyperplane.

[0069] Optionally, the one-to-many target vectors, many-to-one target vectors and many-to-many target vectors corresponding to the basic information and business usage information, the basic information and the basic information of the activated business, and the business usage information and the basic information of the activated business of any two users can be introduced into the target hyperplane. Specifically, each target vector and the head entity object and tail entity object corresponding to each target vector are projected onto the corresponding relational hyperplane, so that the same entity object has different representations in different relational planes in the target hyperplane.

[0070] S403 : Determine the similarity between target data information of any two users among the multiple users based on the target hyperplane and the optimal transmission algorithm.

[0071] It can be understood that the similarity between target data information of any two users among the multiple users can be determined based on the target hyperplane and the optimal transmission algorithm.

[0072] Optionally, the distance and similarity between different entity objects (ie, target data information of any two users among multiple users) can be determined by an optimal transmission algorithm based on different representations of the same entity object on different relationship planes in the target hyperplane.

[0073] In one design, Figure 6 As shown, in a service recommendation method provided in an embodiment of the present application, the method in step S203 specifically includes S501-S502:

[0074] S501: Based on the similarity between target data information of any two users among a plurality of users, determine the correlation between any one of a plurality of feature tags corresponding to one of the two users and a first service activated by the other user.

[0075] S502: Determine a weight value corresponding to any feature tag and the first service according to the degree of association between any feature tag and the first service.

[0076] It can be understood that based on the similarity between the target data information of any two users among multiple users, the correlation between any one of the multiple feature tags corresponding to one of any two users and the first service opened by the other user can be determined, and based on the correlation between any one feature tag and the first service, the corresponding weight value between any one feature tag and the first service can be determined.

[0077] The embodiment of the present application provides a service recommendation method. The present application proposes to introduce a user portrait model and a knowledge graph based on a collaborative filtering algorithm to form a communication user prediction and recommendation method. The user portrait is realized by clustering users, and the knowledge graph is introduced to mine the usage preferences of the target users, thereby predicting and completing the optimal match between the communication user and the package. Specifically, feature labels are established by combining user basic information and user status information (i.e., user service usage information) to construct a user-status user portrait model (i.e., user status label); feature labels are established by combining user basic information and user existing package information (i.e., basic information of the user's activated services) to construct a user-package user portrait model (i.e., service value label and service function label). At the same time, a generative adversarial network is introduced into the knowledge graph to obtain deeper user preferences and similarity results, thereby improving the performance of the knowledge graph embedding model. Furthermore, based on the collaborative filtering algorithm, the user portrait model and the knowledge graph fusion results (i.e., deeper user preferences and similarity results) are used to predict and recommend communication users. Python code can also be used to generate a user portrait data dashboard for data visualization.

[0078] The above mainly introduces the solution provided by the embodiment of the present application from the perspective of method. In order to realize the above functions, it includes hardware structures and / or software modules corresponding to the execution of each function. Those skilled in the art should easily appreciate that, in combination with the units and algorithm steps of each example described in the embodiments disclosed herein, the embodiments of the present application can be implemented in the form of hardware or a combination of hardware and computer software. Whether a function is executed in a hardware or computer software driven hardware manner depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0079] In the embodiment of the present application, a business recommendation method can be divided into functional modules according to the above method example. For example, each functional module can be divided according to each function, or two or more functions can be integrated into one processing module. The above integrated modules can be implemented in the form of hardware or in the form of software functional modules. Optionally, the division of modules in the embodiment of the present application is schematic and is only a logical functional division. In actual implementation, there may be other division methods.

[0080] Figure 7 This is a structural diagram of a business recommendation device provided in an embodiment of the present application. Figure 7 As shown, a service recommendation device 40 is used to improve the accuracy and efficiency of recommending services to users, for example, Figure 2 A service recommendation method is shown. The service recommendation device 40 includes: a processing unit 401 and a determination unit 402;

[0081] Processing unit 401 is configured to perform clustering processing on the multiple users using a preset algorithm and multiple preset rules based on target data information of each user, to obtain at least one feature tag corresponding to each user in the multiple users, wherein the target data information includes at least one of the following: basic information, service usage information, and basic information of activated services; each preset rule corresponds to one feature tag, each feature tag corresponds to at least one user, and the feature tags include at least one of the following: a service value tag, a service function tag, and a user status tag;

[0082] A determining unit 402 is configured to determine, based on the target data information of each user among the multiple users, the similarity between the target data information of any two users using a preset knowledge graph model, where the knowledge graph model is configured to associate the multiple target data information;

[0083] The determining unit 402 is further configured to determine, based on the similarity between target data information of any two users, at least one weight value corresponding to each of the at least one feature tag corresponding to each of the multiple users, the weight value being used to indicate a demand of the user corresponding to the feature tag for one of the multiple preset services;

[0084] The determining unit 402 is further configured to determine, from the multiple services, a target service that is most in demand by any user according to a weight value corresponding to each feature tag in at least one feature tag corresponding to any user among the multiple users.

[0085] In one possible implementation, the service recommendation device 40 further includes an acquisition unit 403; the acquisition unit 403 is used to obtain raw data information of each of the multiple users; the processing unit 401 is further used to perform data preprocessing on the raw data information of each of the multiple users to obtain target data information of each user, where the data preprocessing includes: data cleaning, data error correction, and data unification.

[0086] In one possible implementation, the processing unit 401 is further used to construct a target vector corresponding to the target data information of any two users through a knowledge graph model based on the target data information of each user among multiple users; the processing unit 401 is further used to introduce the target vector corresponding to the target data information of any two users into a target hyperplane, and represent the target data information of each user in multiple relationship planes in the target hyperplane, where the target hyperplane is constructed based on multiple users, multiple services, and user status labels, and the user status labels include at least one of the following: users to be lost, stable users, and users with service changes; the determination unit 402 is further used to determine the similarity between the target data information of any two users among multiple users based on the target hyperplane and the optimal transmission algorithm.

[0087] In one possible implementation, the determination unit 402 is further configured to determine, based on the similarity between the target data information of any two users among the multiple users, the degree of association between any one of the multiple feature tags corresponding to one of the two users and the first service activated by the other user; the determination unit 402 is further configured to determine, based on the degree of association between any one of the multiple feature tags and the first service, the corresponding weight value between any one of the feature tags and the first service.

[0088] In the case of implementing the functions of the above-mentioned integrated modules in the form of hardware, the embodiment of the present application provides a possible structural diagram of the electronic device involved in the above-mentioned embodiment. Figure 8 As shown, an electronic device 60 is used to improve the accuracy and efficiency of recommending services to users, for example, Figure 2 The electronic device 60 includes a processor 601, a memory 602, and a bus 603. The processor 601 and the memory 602 may be connected via the bus 603.

[0089] Processor 601 is the control center of the communication device and can be a single processor or a collective term for multiple processing elements. For example, processor 601 can be a general-purpose central processing unit (CPU) or other general-purpose processor. The general-purpose processor can be a microprocessor or any conventional processor.

[0090] As an embodiment, the processor 601 may include one or more CPUs, such as Figure 8 CPU 0 and CPU 1 are shown in Figure 1.

[0091] The memory 602 may be a read-only memory (ROM) or other type of static storage device that can store static information and instructions, a random access memory (RAM) or other type of dynamic storage device that can store information and instructions, an electrically erasable programmable read-only memory (EEPROM), a magnetic disk storage medium or other magnetic storage device, or any other medium that can be used to carry or store desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto.

[0092] As a possible implementation, memory 602 can exist independently of processor 601 and can be connected to processor 601 via bus 603 to store instructions or program codes. When processor 601 calls and executes the instructions or program codes stored in memory 602, a service recommendation method provided in an embodiment of the present application can be implemented.

[0093] In another possible implementation, the memory 602 may also be integrated with the processor 601 .

[0094] Bus 603 can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus. This bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 8 Only one thick line is used in the diagram, but this does not mean that there is only one bus or one type of bus.

[0095] It should be pointed out that Figure 8 The structure shown does not constitute a limitation on the electronic device 60. Figure 8 In addition to the components shown, the electronic device 60 may include more or fewer components than shown, or combine certain components, or arrange the components differently.

[0096] As an example, combining Figure 7 The functions implemented by the processing unit 401, the determining unit 402 and the obtaining unit 403 in the business recommendation device 40 are similar to those implemented by Figure 8 The functions of the processor 601 in are the same.

[0097] Optional, such as Figure 8 As shown, the electronic device 60 provided in the embodiment of the present application may further include a communication interface 604 .

[0098] The communication interface 604 is used to connect to other devices via a communication network. The communication network can be Ethernet, wireless access network, wireless local area network (WLAN), etc. The communication interface 604 can include a receiving unit for receiving data and a sending unit for sending data.

[0099] In one design, in the electronic device provided in the embodiment of the present application, the communication interface can also be integrated into the processor.

[0100] Through the description of the above embodiments, those skilled in the art will clearly understand that for the sake of convenience and brevity, only the division of the above-mentioned functional units is used as an example. In actual applications, the above-mentioned functions can be distributed and completed by different functional units as needed, that is, the internal structure of the device can be divided into different functional units to complete all or part of the functions described above. The specific working processes of the above-mentioned systems, devices, and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0101] An embodiment of the present application further provides a computer-readable storage medium, in which instructions are stored. When a computer executes the instructions, the computer executes each step in the method flow shown in the above method embodiment.

[0102] An embodiment of the present application provides a computer program product comprising instructions. When the instructions are executed on a computer, the computer is caused to execute a service recommendation method in the above method embodiment.

[0103] Among them, the computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination thereof. More specific examples of computer-readable storage media (a non-exhaustive list) include: an electrical connection with one or more wires, a portable computer disk, a hard disk. Random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), registers, hard disks, optical fibers, portable compact disc read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any other form of computer-readable storage media in a suitable combination of the above, or values ​​in this field.

[0104] An exemplary storage medium is coupled to a processor so that the processor can read information from the storage medium and write information to the storage medium. Of course, the storage medium can also be an integral part of the processor. The processor and storage medium can be located in an application-specific integrated circuit (ASIC).

[0105] In the embodiments of the present application, a computer-readable storage medium may be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0106] Since the electronic device, computer-readable storage medium, and computer program product in the embodiments of the present application can be applied to the above method, the technical effects that can be obtained can also refer to the above method embodiments, and the embodiments of the present application will not be repeated here.

[0107] The above are only specific implementation methods of the present application, but the protection scope of the present application is not limited thereto. Any changes or replacements within the technical scope disclosed in the present application should be included in the protection scope of the present application.

Claims

1. A business recommendation method, characterized in that: The method comprises: Based on target data information of each user among a plurality of users, clustering the plurality of users using a preset algorithm and a plurality of preset rules to obtain at least one feature tag corresponding to each of the plurality of users, the target data information including at least one of the following: basic information, service usage information, and basic information of activated services, one preset rule corresponding to one feature tag, one feature tag corresponding to at least one user, and the feature tag including at least one of the following: a service value tag, a service function tag, and a user status tag; Based on the target data information of each user among the multiple users, determining the similarity between the target data information of any two users through a preset knowledge graph model, wherein the knowledge graph model is used to associate the multiple target data information; Determining, based on the similarity between target data information of any two users, at least one weight value corresponding to each feature tag of at least one feature tag corresponding to each of the multiple users, the weight value being used to indicate a demand of the user corresponding to the feature tag for one of the multiple preset services; The determining, based on the similarity between the target data information of any two users, at least one weight value corresponding to each feature tag of at least one feature tag corresponding to each user of the plurality of users includes: determining, based on the similarity between the target data information of any two users among the multiple users, a correlation between any one of the multiple feature tags corresponding to one of the two users and a first service activated by the other user; Determining a weight value corresponding to each of the feature tags and the first service according to a degree of association between the feature tags and the first service; According to the weight value corresponding to each feature tag in at least one feature tag corresponding to any one of the multiple users, a target service with the greatest demand from the any one of the multiple users is determined from the multiple services.

2. The method according to claim 1, characterized in that The method further comprises: Obtaining original data information of each user among the multiple users; Data preprocessing is performed on the original data information of each user among the multiple users to obtain target data information of each user, and the data preprocessing includes: data cleaning, data error correction, and data unification.

3. The method according to claim 1 or 2, characterized in that The determining, based on the target data information of each user among the multiple users, the similarity between the target data information of any two users by using a preset knowledge graph model includes: Based on the target data information of each user among the multiple users, constructing a target vector corresponding to the target data information of any two users through the knowledge graph model; Introducing the target vector corresponding to the target data information of any two users into a target hyperplane, and representing the target data information of each user in multiple relationship planes in the target hyperplane, wherein the target hyperplane is constructed based on the multiple users, the multiple services, and the user status labels, wherein the user status label includes at least one of the following: a user to be lost, a stable user, and a user who changes services; Based on the target hyperplane and the optimal transmission algorithm, similarity between target data information of any two users among the multiple users is determined.

4. A business recommendation device, characterized in that: The service recommendation device includes: a processing unit and a determination unit; The processing unit is configured to perform clustering processing on the multiple users using a preset algorithm and multiple preset rules based on target data information of each user to obtain at least one feature tag corresponding to each user in the multiple users, wherein the target data information includes at least one of the following: basic information, service usage information, and basic information of activated services; one preset rule corresponds to one feature tag, one feature tag corresponds to at least one user, and the feature tag includes at least one of the following: a service value tag, a service function tag, and a user status tag; The determining unit is configured to determine, based on the target data information of each user among the multiple users, the similarity between the target data information of any two users through a preset knowledge graph model, wherein the knowledge graph model is configured to associate the multiple target data information; The determining unit is further configured to determine, based on the similarity between the target data information of any two users, at least one weight value corresponding to each of the at least one feature tag corresponding to each of the multiple users, the weight value being used to indicate a degree of demand of the user corresponding to the feature tag for one of the multiple preset services; The determining unit is further configured to determine, from the multiple services, a target service that is most in demand by any one of the multiple users based on a weight value corresponding to each feature tag in at least one feature tag corresponding to any one of the multiple users; The determining unit is further configured to determine, based on the similarity between the target data information of any two users among the multiple users, a correlation between any one of the multiple feature tags corresponding to one of the two users and the first service activated by the other user; The determining unit is further configured to determine a weight value corresponding to any one of the feature tags and the first service according to a degree of association between any one of the feature tags and the first service.

5. The business recommendation device according to claim 4, characterized in that: The business recommendation device further includes an acquisition unit; The acquiring unit is configured to acquire original data information of each of the multiple users; The processing unit is further configured to perform data preprocessing on the original data information of each user among the multiple users to obtain target data information of each user, wherein the data preprocessing includes: data cleaning, data error correction, and data unification.

6. The business recommendation device according to claim 4 or 5, characterized in that: The processing unit is further configured to construct, based on the target data information of each user among the multiple users, a target vector corresponding to the target data information of any two users through the knowledge graph model; The processing unit is further configured to introduce a target vector corresponding to target data information of any two users into a target hyperplane, and represent the target data information of each user in multiple relationship planes in the target hyperplane, wherein the target hyperplane is constructed based on the multiple users, the multiple services, and the user status labels, wherein the user status label includes at least one of the following: a user to be lost, a stable user, and a user with a service change; The determining unit is further configured to determine the similarity between target data information of any two users among the multiple users based on the target hyperplane and the optimal transmission algorithm.

7. An electronic device, characterized in that: include: A processor and a memory; wherein the memory is used to store one or more programs, and the one or more programs include computer-executable instructions. When the electronic device is running, the processor executes the computer-executable instructions stored in the memory to enable the electronic device to execute a business recommendation method according to any one of claims 1 to 3.

8. A computer-readable storage medium storing one or more programs, characterized in that: The one or more programs include instructions, which, when executed by a computer, enable the computer to perform a service recommendation method according to any one of claims 1 to 3.

Images