CN115293452A

CN115293452A - User behavior prediction method and device, computer equipment and storage medium

Info

Publication number: CN115293452A
Application number: CN202211025093.4A
Authority: CN
Inventors: 褚晓晶
Original assignee: Industrial and Commercial Bank of China Ltd ICBC
Current assignee: Industrial and Commercial Bank of China Ltd ICBC
Priority date: 2022-08-25
Filing date: 2022-08-25
Publication date: 2022-11-04

Abstract

The application relates to a user behavior prediction method, a user behavior prediction device, computer equipment and a storage medium, and relates to the field of artificial intelligence. The method comprises the following steps: acquiring report query behavior data of a user to be predicted; acquiring service behavior data of a service user; aggregating the report query behavior data and the service behavior data to obtain user behavior theme data; and inputting the user behavior theme data into a pre-trained behavior prediction model to obtain a behavior prediction result corresponding to the user to be predicted. By adopting the method, the user behavior prediction can be realized when the resource conversion data between the user and the financial enterprise is lacked.

Description

User behavior prediction method and device, computer equipment and storage medium

Technical Field

The present application relates to the field of artificial intelligence technologies, and in particular, to a method, an apparatus, a computer device, a storage medium, and a computer program product for predicting user behavior.

Background

The financial enterprise accumulates massive high-value data in the process of developing business, wherein the high-value data comprise user basic information and resource conversion data between the user and the financial enterprise. And pushing the product for the user by predicting the next action of the user so as to prevent the user from losing. The traditional approach is to make behavioral predictions for a user based on resource transformation data between the user and the financial enterprise.

However, conventional approaches fail to predict user behavior in the absence of resource translation data between the user and the financial enterprise. Therefore, how to realize the user behavior prediction in the absence of resource conversion data between the user and the financial enterprise becomes a technical problem to be solved at present.

Disclosure of Invention

In view of the foregoing, it is desirable to provide a user behavior prediction method, apparatus, computer device, computer readable storage medium, and computer program product, which can realize user behavior prediction in the absence of resource transformation data between a user and a financial enterprise.

In a first aspect, the present application provides a method for predicting user behavior. The method comprises the following steps:

acquiring report query behavior data of a user to be predicted;

acquiring service behavior data of a service user;

aggregating the report query behavior data and the service behavior data to obtain user behavior theme data;

and inputting the user behavior theme data into a pre-trained behavior prediction model to obtain a behavior prediction result corresponding to the user to be predicted.

In one embodiment, the obtaining of the report query behavior data of the user to be predicted comprises:

acquiring report query behavior data of a user to be predicted from a plurality of self-service channels; the self-service channels comprise preset application programs and self-service business equipment.

In one embodiment, aggregating the report query behavior data and the service behavior data to obtain user behavior topic data includes:

performing data conversion on the report query behavior data and the service behavior data through a big data platform to obtain effective query behavior data and effective service behavior data;

and extracting resource data from the effective query behavior data and the effective service behavior data, and determining the resource data as user behavior subject data.

In one embodiment, inputting the user behavior topic data into a pre-trained behavior prediction model, and obtaining a behavior prediction result corresponding to the user to be predicted includes:

inputting the user behavior theme data into a pre-trained behavior prediction model, and performing prediction operation on the user behavior theme data through the behavior prediction model to obtain the relevance between the report inquiry behavior data and the service behavior data;

and predicting a behavior prediction result corresponding to the user to be predicted according to the relevance through the behavior prediction model.

In one embodiment, the method further includes:

determining the category of a target product according to the behavior prediction result;

determining a target product corresponding to the target product category in preset category products;

and pushing the target product to the user to be predicted.

In one embodiment, before the inputting the user behavior topic data into the pre-trained behavior prediction model, the method further includes:

classifying the user to be predicted according to the user behavior theme data to obtain a user category corresponding to the user to be predicted;

determining a target user in the users to be predicted according to the user category and a preset screening condition;

extracting target subject data corresponding to the target user from the user behavior subject data;

the step of inputting the user behavior theme data into a pre-trained behavior prediction model to obtain a behavior prediction result corresponding to the user to be predicted comprises the following steps:

and inputting the target theme data into a pre-trained behavior prediction model to obtain a behavior prediction result corresponding to the target user.

In a second aspect, the present application further provides a user behavior prediction apparatus. The device comprises:

the query data acquisition module is used for acquiring report query behavior data of a user to be predicted;

the service data acquisition module is used for acquiring service behavior data of a service user;

the data aggregation module is used for aggregating the report query behavior data and the service behavior data to obtain user behavior theme data;

and the behavior prediction module is used for inputting the user behavior theme data into a pre-trained behavior prediction model to obtain a behavior prediction result corresponding to the user to be predicted.

In a third aspect, the present application also provides a computer device. The computer device comprises a memory storing a computer program and a processor implementing the following steps when executing the computer program:

acquiring report query behavior data of a user to be predicted;

acquiring service behavior data of a service user;

In a fourth aspect, the present application further provides a computer-readable storage medium. The computer-readable storage medium having stored thereon a computer program which, when executed by a processor, performs the steps of:

acquiring report query behavior data of a user to be predicted;

acquiring service behavior data of a service user;

In a fifth aspect, the present application further provides a computer program product. The computer program product comprising a computer program which when executed by a processor performs the steps of:

acquiring report query behavior data of a user to be predicted;

acquiring service behavior data of a service user;

The user behavior prediction method, the user behavior prediction device, the computer equipment, the storage medium and the computer program product are used for acquiring the report inquiry behavior data of the user to be predicted, acquiring the service behavior data of the service user, performing aggregation processing on the report inquiry behavior data and the service behavior data, and inputting the user behavior subject data obtained through aggregation into a pre-trained behavior prediction model to obtain the behavior prediction result corresponding to the user to be predicted. By acquiring the report query behavior data of the user, the behavior prediction data of potential users and low-end users are enriched, and the behavior prediction of the user can be performed under the condition of lacking of resource conversion data. The report query behavior data and the service behavior data are aggregated, so that the implicit value of the data is mined, and then the user behavior is predicted through the behavior prediction model, so that the accuracy of behavior prediction is improved.

Drawings

FIG. 1 is a diagram of an application environment of a method for predicting user behavior in one embodiment;

FIG. 2 is a flow diagram that illustrates a methodology for predicting user behavior in one embodiment;

FIG. 3 is a flowchart illustrating a step of aggregating report query behavior data and service behavior data to obtain user behavior topic data in one embodiment;

FIG. 4 is a schematic diagram illustrating a data aggregation step performed by a big data platform according to another embodiment;

FIG. 5 is a block diagram of an architectural model of a big data platform according to another embodiment;

FIG. 6 is a flowchart illustrating a method for predicting user behavior in accordance with another embodiment;

FIG. 7 is a flowchart illustrating a method for predicting user behavior in accordance with another embodiment;

FIG. 8 is a block diagram showing the structure of a user behavior prediction apparatus according to an embodiment;

FIG. 9 is a diagram illustrating an internal structure of a computer device according to an embodiment.

Detailed Description

In order to make the objects, technical solutions and advantages of the present application more apparent, the present application is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of and not restrictive on the broad application.

In the process of carrying out business, the financial enterprise pushes products to the user by predicting the next action of the user so as to prevent the user from losing. The traditional method is used for predicting the user behavior based on resource conversion data between the user and the financial enterprise, but for the user lacking the resource conversion data, the traditional method cannot predict the user behavior.

In order to solve the technical problem, a user behavior prediction method is provided.

The user behavior prediction method provided by the embodiment of the application can be applied to the application environment shown in fig. 1. Wherein the data collection device 102 communicates with the server 104 via a network. The data acquisition device can be a plurality of devices. The data collection device 102 collects report query behavior data of a user to be predicted, and sends the collected data to the server 104. The server 104 acquires the business behavior data of the business user, aggregates the report query behavior data and the business behavior data to obtain user behavior topic data, and inputs the user behavior topic data into a pre-trained behavior prediction model to obtain a behavior prediction result corresponding to the user to be predicted. The data collection device 102 may be, but is not limited to, various personal computers, notebook computers, smart phones, and tablet computers. The server 104 may be implemented as a stand-alone server or a server cluster comprised of multiple servers.

In one embodiment, as shown in fig. 2, a user behavior prediction method is provided, which is described by taking the method as an example applied to the server in fig. 1, and includes the following steps:

step 202, report query behavior data of the user to be predicted is obtained.

The user to be predicted refers to a user with a report query behavior. The report query behavior data refers to a query record generated by report query behavior of a user to be predicted.

Specifically, the server obtains report query behavior data collected by the data collection device. The user can freely inquire the credit report of the user through a plurality of preset self-service channels so as to know the credit condition of the user. Therefore, the query records generated by the report query behavior of the user to be predicted can be collected through the data collection devices corresponding to the auxiliary channels respectively to serve as report query behavior data. The plurality of self-service channels can be obtained from preset application programs and self-service equipment. The report query behavior data may include query time, query mode, location area to which the user belongs, user basic information, and other data.

Step 204, acquiring service behavior data of the service user.

The service user refers to a user transacting a service in a financial institution. The service behavior data refers to service data related to service users.

The server can also obtain business behavior data from various business data sources such as a business application system, a business database, a business application log and the like, wherein the business behavior data can comprise user basic information, resource conversion data between the user and a financial enterprise, user credit score data, contract data, resource conversion channel data and external market data. For example, the transaction may include a loan transaction, a financial transaction, and the like. The resource conversion data may be transaction data. The external market data may include market regulatory policies, such as housing loan interest rate policies.

And step 206, aggregating the report query behavior data and the service behavior data to obtain user behavior theme data.

The aggregation processing refers to integrating and extracting scattered report query behavior data and business behavior data. The user behavior theme data refers to aggregated data taking user behaviors as themes.

The server integrates the acquired report query behavior data and the acquired business behavior data, and specifically, the report query behavior data and the business behavior data can be divided into relational data and non-relational data, and the relational data and the non-relational data are integrated respectively to obtain integrated data. Relational data refers to data that is structured, such as data that can be used to generate tables. Non-relational data refers to unstructured data, such as textual information. And extracting resource data from the integrated data as user behavior topic data. The resource data refers to data assets, and may include user characteristic data and statistical index data.

Alternatively, the user behavior topic data may be a user behavior topic aggregation table.

And 208, inputting the user behavior theme data into a pre-trained behavior prediction model to obtain a behavior prediction result corresponding to the user to be predicted.

The behavior prediction result refers to the probability that the user to be predicted purchases products of each product category.

The server is stored with a behavior prediction model in advance, and the behavior prediction model is obtained by training a large amount of user report query behavior data and user service behavior data. The behavior prediction model may be an existing model for behavior prediction. For example, the behavior prediction model may include, but is not limited to, BERT (pre-trained language characterization model), convolutional Neural network model (CNN), and Random Forest model (RF).

The server calls a pre-trained behavior prediction model, user theme behavior data are input into the behavior prediction model, and the probability that the user to be predicted purchases products of each product category is predicted according to the user behavior theme data through the behavior prediction model. For example, product categories may include funds, insurance, and deposits. The behavior prediction result corresponding to the user to be predicted may include a probability of purchasing a fund product of 85%, a probability of purchasing an insurance product of 70%, and a probability of purchasing a deposit product of 50%.

Furthermore, the server can screen target users according to the user behavior theme data, wherein the target users refer to potential users who have not transacted business in the financial enterprise, and therefore customer obtaining channels are increased.

The user behavior prediction method comprises the steps of obtaining report inquiry behavior data of a user to be predicted, obtaining service behavior data of a service user, carrying out aggregation processing on the report inquiry behavior data and the service behavior data, inputting user behavior subject data obtained through aggregation into a pre-trained behavior prediction model, and obtaining a behavior prediction result corresponding to the user to be predicted. By acquiring the report query behavior data of the user, the behavior prediction data of potential users and low-end users are enriched, and the behavior prediction of the user can be performed under the condition of lacking of resource conversion data. The report query behavior data and the service behavior data are aggregated, so that the implicit value of the data is mined, and then the user behavior is predicted through the behavior prediction model, so that the accuracy of behavior prediction is improved.

In one embodiment, obtaining report query behavior data of a user to be predicted comprises: acquiring report query behavior data of a user to be predicted from a plurality of self-service channels; the plurality of self-service channels comprise preset application programs and self-service equipment.

The server can acquire the report query behavior data acquired by the data acquisition equipment from a plurality of channels such as a preset application program and self-service equipment. Reporting query behavior data may also be referred to as query logs or query logs. For example, the preset application program may be a personal internet bank, a personal cell phone bank, or the like. The self-service device may be a website intelligent teller machine.

Specifically, by embedding a data collection SDK (Software Development Kit) in a preset application program, when a user queries a report through a prediction application program, the SDK can record each query log of the user in real time, so that the data collection device obtains the query logs of the user collected by the SDK. Meanwhile, the query log on the user self-service equipment can be acquired through the data acquisition equipment corresponding to the self-service equipment.

In the embodiment, the report query behavior data of the user to be predicted is acquired from a plurality of channels such as the preset application program and the self-service device, so that the integrity and comprehensiveness of the report query behavior data are improved, and the behavior prediction accuracy is improved.

In an embodiment, as shown in fig. 3, aggregating the report query behavior data and the service behavior data to obtain the user behavior topic data includes:

step 302, data conversion is performed on the report query behavior data and the service behavior data through the big data platform, so as to obtain effective query behavior data and effective service behavior data.

And step 304, extracting resource data from the effective query behavior data and the effective business behavior data, and determining the resource data as user behavior theme data.

The report query behavior data may include query time, query mode, location area to which the user belongs, user basic information, and other data. The business behavior data may include user basic information, resource conversion data between the user and the financial enterprise, user credit score data, contract data, resource conversion channel data, and external market data. For example, the transaction may include a loan transaction, a financial transaction, and the like. The resource conversion data may be transaction data.

The server can send the report query behavior data and the service behavior data to the big data platform, and data aggregation is carried out through the big data platform. A schematic diagram of the data aggregation steps through a big data platform can be shown in fig. 4. The big data platform can be a Hadoop-based big data platform. The customer basic information, the transaction information, the grading information, the contract information, the transaction channel information and the external market information in the data source respectively represent the user basic information, the resource conversion data between the user and the financial enterprise, the user credit grading data, the contract data, the resource conversion channel data and the external market data in the business behavior data. Client credit report query activity refers to reporting query activity data. The information infrastructure of a big Data platform includes an Extract-Transform-Load (ETL), an ODS (Operational Data Store), a Data warehouse, and a Data mart.

After the big data platform obtains data of a plurality of data sources of report query behavior data and business behavior data, a data extraction process is started through an ETL tool, the report query behavior data and the business behavior data are extracted to an intermediate layer, then a data conversion process is started, data conversion is carried out on the data of the intermediate layer, and the data conversion process can comprise cleaning, conversion and loading. Because the acquired report query behavior data and the acquired service behavior data are both original data, redundancy, conflict, abnormality and deficiency exist, and data conversion needs to be carried out through a big data platform. Specifically, the data conversion process may include performing data cleaning on the acquired report query behavior data and the acquired business behavior data through an ETL tool, and the data cleaning manner may include uniform data expression format, sorting data, screening duplicate data, merging or dividing data items, invalid data deletion, missing column deletion, missing value replacement, abnormal value processing, row deduplication, column deduplication, code replacement, data filtering, type conversion, and format conversion. And storing the cleaned data into the ODS. And then, carrying out data conversion on the data cleaned in the ODS through an ETL tool to obtain effective query behavior data and effective service behavior data. The data conversion mode may include performing data conversion on the cleaned data according to a preset conversion rule. For example, the preset conversion rules may include adding sequences, adding constants, line-column conversion, merging records, data item splicing, data item splitting, data type conversion, character string replacement, character string padding, character string clipping, character string truncation, numerical extraction, numerical padding, value mapping, computation function conversion, script execution, data set splitting, data set merging, data set joining, data set sorting, and the like. The valid query behavior data and the valid business behavior data are stored in a data repository, i.e., the business data repository in fig. 4. The business data warehouse can perform data calculation according to the model and the database and the data calculation flow, and can also perform data summarization according to the model and the database and the data summarization flow. And then starting a converted data loading process through an ETL tool, loading effective query behavior data and effective business behavior data in a data warehouse, and performing relational data integration and non-relational data integration on the effective query behavior data and the effective business behavior data, so as to extract resource data from the integrated data to be used as user behavior theme data. The resource data refers to data assets, and may include user characteristic data and general statistical indicator data with sharing reusability. The user characteristic data may include user profile characteristics such as the user's age, income level, occupation, etc., or may also include product attribute characteristics such as product type, product price, etc. The general statistical indicator data may indicate data categories such as user basic information indicators, resource conversion indicators, credit score indicators, and the like. Storing the user behavior theme data in a data warehouse, and also storing the user behavior theme data in a data mart.

Illustratively, the architectural model of a big data platform may be as shown in FIG. 5. The data sources of the transaction, that is, the data sources of the resource conversion data between the user and the financial enterprise in the business behavior data, may include logs and distributed application logs. The sources of informatization, i.e. sources of Data other than the resource transformation Data in the business behavior Data, may include DWH (Data repository) databases, other databases, transaction systems, and the like. The newly added external information sources, namely data sources for reporting the query behavior data, can comprise social networks, networks and the like.

In the embodiment, data aggregation is performed through a big data platform, so that high sharing and integration of data are realized, the implied value of the data is mined, and the accuracy of user behavior theme data is improved.

In one embodiment, inputting the user behavior topic data into a pre-trained behavior prediction model, and obtaining a behavior prediction result corresponding to a user to be predicted includes: inputting user behavior theme data into a pre-trained behavior prediction model, and performing prediction operation on the user behavior theme data through the behavior prediction model to obtain the relevance between report inquiry behavior data and service behavior data; and predicting a behavior prediction result corresponding to the user to be predicted according to the relevance through a behavior prediction model.

The service behavior data refers to service data related to service users.

A plurality of important factors of business behavior data are deployed in the pre-trained behavior prediction model. For example, the behavior prediction model may be any one of a BERT model, a convolutional neural network model, and a random forest model.

Specifically, by means of a behavior prediction model, concurrent calculation is performed on important factors of user characteristic data and business behavior data in user behavior theme data by means of multiple threads, and the relevance between report query behavior data and business behavior data is obtained, so that the probability that a user to be predicted purchases products of each product category is determined according to the calculated relevance between the report query behavior data and the business behavior data. Wherein the greater the value of the association, the higher the probability of purchasing a product of the corresponding product category.

In this embodiment, the server can effectively analyze the relevance between the report query behavior data and the business behavior data according to the user behavior topic data. Meanwhile, the prediction efficiency of the behavior prediction result can be improved through multi-thread concurrent computation.

In one embodiment, the method further comprises: determining the target product type according to the behavior prediction result; determining a target product corresponding to the target product category in preset category products; and pushing the target product to a user to be predicted.

The target product category refers to a product category needing to be pushed. The preset category product refers to the existing category product of the financial institution. The target product refers to a business product needing to be pushed. Such as fund products, insurance products, deposit products, and the like.

The behavior prediction result includes the probability of purchasing a product of each product category by the user to be predicted, and the server can select the product category with the highest probability from the multiple probabilities as the target product category. The preset category product comprises a plurality of product categories. The server can search the target product category in the preset category products, obtain the target product corresponding to the target product category, and then push the target product to the user to be predicted.

In this embodiment, a target product category is selected from a plurality of product categories of the behavior prediction result, a target product corresponding to the target product category is determined in a preset category product, and pushing is performed, so that the pushing range can be further narrowed, the product pushing is more accurate, the user requirements can be better met, and the user conversion rate and the user retention rate can be improved.

In one embodiment, before inputting the user behavior topic data into the pre-trained behavior prediction model, the method further comprises: classifying users to be predicted according to the user behavior theme data to obtain user categories corresponding to the users to be predicted; determining a target user in the users to be predicted according to the user category and a preset screening condition; extracting target subject data corresponding to a target user from the user behavior subject data; inputting the user behavior theme data into a pre-trained behavior prediction model, and obtaining a behavior prediction result corresponding to a user to be predicted comprises the following steps: and inputting the target subject data into a pre-trained behavior prediction model to obtain a behavior prediction result corresponding to the target user.

User characteristic data is included in the user behavior theme data, and the user characteristic data can include user portrait characteristics such as the age, income level, occupation, location area where the user is located and the like of the user, or can also include product attribute characteristics such as product type, product price and the like. The server can classify the users to be predicted according to the user characteristic data in the user behavior theme data, and classify the users to be predicted containing the same characteristic data into a user category. Therefore, the user category corresponding to the user to be predicted may be one or more. For example, user categories may include 20-30 years of age, moderate income level, free occupation, living in Beijing, and the like. And determining a target user in the users to be predicted according to the user category and the preset screening condition. Wherein the predetermined screening condition may be an age of 25-35 years, income level and stable occupation. The server can determine users meeting preset screening conditions according to the user categories as target users.

After the target user is obtained, extracting target subject data corresponding to the target user from the user behavior subject data, and inputting the target subject data into a pre-trained behavior prediction model to obtain a behavior prediction result corresponding to the target user. The behavior prediction mode of the target user is the same as that of the user to be predicted, and details are not repeated here.

In the embodiment, the users to be predicted are classified according to the user behavior theme data to obtain the user categories corresponding to the users to be predicted, the target users are further determined, the target users are pushed, real target users can be screened from a large number of users to be predicted, and the product pushing accuracy can be improved.

In one embodiment, the method further comprises: and sending the user behavior theme data to a data analysis terminal to instruct the data analysis terminal to carry out risk prediction on the user to be predicted according to the user behavior theme data.

Specifically, the data analysis terminal may obtain the query times of the report query behavior data in the user behavior topic data, and when the query times meet a preset condition, it indicates that the user to be predicted is in an abnormal state. The preset condition may be that the number of queries within a preset time period is exceeded. For example, the number of queries exceeds 5 times a day. The data sources of the user credit risk assessment are increased by analyzing the report query behavior data, and the monitoring of the abnormal query behavior of the user is realized, so that the attention to the user credit risk is strengthened.

Furthermore, the data analysis terminal can also determine users who have transacted business, such as users in post-loan management, according to the user behavior topic data. And performing credit verification on the transacted user to predict the risk early warning level of the user. Specifically, each risk early warning level may correspond to one credit assessment score data interval. The credit verification may be to acquire user credit score data of the handled service, compare the user credit score data with a plurality of preset risk early warning levels for credit assessment score intervals, and determine the risk early warning level as a risk early warning level corresponding to the user having handled the service when the user credit score data falls into the credit assessment score interval corresponding to the risk early warning level. Therefore, the server can generate early warning prompt information according to the risk early warning level and send the early warning prompt information to the associated terminal, and the associated terminal can take risk processing measures for corresponding users according to the received early warning prompt information so as to avoid the influence of overlarge user risk on financial enterprises.

In one embodiment, when the report query behavior data shows no report, it indicates that the credit data of the user cannot be acquired, and the report query behavior data is useless for user behavior prediction at this time, but on the other hand, a data analyzer corresponding to the data analysis terminal knows that the behavior prediction accuracy of the user without the report needs to be improved in other ways.

Exemplarily, as shown in fig. 6, it is a schematic flowchart of a user behavior prediction method in another embodiment. The user behavior prediction method can be applied to a user behavior prediction system, and the user behavior prediction system comprises a technical basic service layer, a business basic service layer and a business product service layer.

1. Processing a client behavior theme aggregation table: in the business basic service layer, big data processing, namely aggregation processing, is carried out on the customer behavior basic data, namely report query behavior data and business behavior data by calling a big data platform in the technical basic service layer, so that a customer behavior topic aggregation table, namely user behavior topic data is obtained, and the customer behavior topic aggregation table is stored in a data warehouse.

2. Query service access customer behavior topic aggregation table: and the user behavior prediction system provides a client behavior topic aggregation table query service for data analysis personnel.

3. Query service access customer behavior base data: the user behavior prediction system provides a customer behavior basic data query service for data analysis personnel.

4. Providing model services and asynchronous query services: the data analysis personnel can log in the user behavior prediction system and provide model service and asynchronous query service for the data analysis personnel through the data analysis service of the user behavior prediction system. In the data query process, the data analysis personnel are subjected to identity verification according to the user authority information, after the identity verification is passed, a client behavior theme aggregation table and client behavior basic data in a data warehouse can be queried, and asynchronous query information is generated according to a query result. And evaluating the abnormal query information by a data analyst to obtain model management information. The data analyst can set corresponding query fields in the query process to obtain report customization information.

5. Providing flexible exploration of business user behavior data, model management and report customization and query: through the data query process in the above 1 to 4, the user behavior data, that is, the user behavior topic data, can be queried. Analyzing and managing the user behavior theme data, including performing behavior prediction on the user behavior theme data to obtain a behavior prediction result corresponding to the user to be predicted, namely behavior prediction information. And risk prediction can be carried out on the user to be predicted according to the user behavior theme data to obtain behavior risk information.

Furthermore, in order to realize the data query function of the user behavior system, the user behavior topic data is required to be utilized to rely on big data such as big data cloud computing, machine learning and deep learning and the technology of the related field of artificial intelligence, various existing recall and sorting algorithm results are fused, barriers among various user behavior data are broken, a user behavior prediction service taking a user as a center is constructed, a data query service is provided for data analysts of a data analysis terminal around the principles of 'business transparency, data regression and service integration', the user behavior topic data and the business behavior data are accessed, the behaviors of the user are explored, and the process of data analysis is realized.

Optionally, a data analyst of the data analysis terminal can flexibly explore data in the user behavior system by means of an existing data visualization tool.

In another embodiment, as shown in fig. 7, there is provided a user behavior prediction method, including the steps of:

step 702, acquiring report query behavior data of a user to be predicted from a plurality of self-service channels; the plurality of self-service channels comprise preset application programs and self-service equipment.

Step 704, obtaining the service behavior data of the service user.

Step 706, performing data conversion on the report query behavior data and the service behavior data through the big data platform to obtain effective query behavior data and effective service behavior data.

And 708, extracting resource data from the effective query behavior data and the effective service behavior data, and determining the resource data as user behavior theme data.

And 710, classifying the users to be predicted according to the user behavior topic data to obtain the user categories corresponding to the users to be predicted.

And 712, determining a target user in the users to be predicted according to the user category and a preset screening condition.

And 714, extracting target subject data corresponding to the target user from the user behavior subject data.

And 716, inputting the target theme data into a pre-trained behavior prediction model to obtain a behavior prediction result corresponding to the target user.

Step 718, determining the target product type according to the behavior prediction result.

And 720, determining a target product corresponding to the target product category in the preset category products.

And step 722, pushing the target product to the user to be predicted.

In the embodiment, by adding a data source of report query behavior data, the behavior prediction data of potential users and low-end users are enriched, and the behavior prediction of the users can be performed under the condition of lacking resource conversion data. Data aggregation is carried out through a big data platform, high sharing and integration of data are achieved, the implied value of the data is mined, and the accuracy of the user behavior theme data is improved. The product pushing accuracy can be improved by screening real target users from a large number of users to be predicted. The target product category is selected from the multiple product categories of the behavior prediction result, the target product corresponding to the target product category is determined in the preset category products, and pushing is carried out, so that the pushing range can be further narrowed, the product pushing is more accurate, the user requirements are met, and the user conversion rate and the user retention rate can be improved.

It should be understood that, although the steps in the flowcharts related to the embodiments as described above are sequentially displayed as indicated by arrows, the steps are not necessarily performed sequentially as indicated by the arrows. The steps are not performed in the exact order shown and described, and may be performed in other orders, unless explicitly stated otherwise. Moreover, at least a part of the steps in the flowcharts related to the embodiments described above may include multiple steps or multiple stages, which are not necessarily performed at the same time, but may be performed at different times, and the execution order of the steps or stages is not necessarily sequential, but may be rotated or alternated with other steps or at least a part of the steps or stages in other steps.

Based on the same inventive concept, the embodiment of the application also provides a user behavior prediction device for implementing the user behavior prediction method. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme described in the above method, so the specific limitations in one or more embodiments of the user behavior prediction device provided below may refer to the limitations on the user behavior prediction method in the above, and details are not repeated here.

In one embodiment, as shown in fig. 8, there is provided a user behavior prediction apparatus including: a query data acquisition module 802, a service data acquisition module 804, a data aggregation module 806, and a behavior prediction module 808, wherein:

a query data obtaining module 802, configured to obtain report query behavior data of a user to be predicted;

a service data obtaining module 804, configured to obtain service behavior data of a service user;

the data aggregation module 806 is configured to aggregate the report query behavior data and the service behavior data to obtain user behavior topic data;

and the behavior prediction module 808 is configured to input the user behavior theme data into a pre-trained behavior prediction model to obtain a behavior prediction result corresponding to the user to be predicted.

In one embodiment, the query data obtaining module 802 is further configured to obtain report query behavior data of the user to be predicted from a plurality of self-service channels; the plurality of self-service channels comprise preset application programs and self-service equipment.

In one embodiment, the data aggregation module 806 is further configured to perform data conversion on the report query behavior data and the service behavior data through the big data platform to obtain effective query behavior data and effective service behavior data; and extracting resource data from the effective query behavior data and the effective service behavior data, and determining the resource data as user behavior theme data.

In one embodiment, the behavior prediction module 808 is further configured to input the user behavior theme data into a pre-trained behavior prediction model, and perform prediction operation on the user behavior theme data through the behavior prediction model to obtain the relevance between the report query behavior data and the service behavior data; and predicting a behavior prediction result corresponding to the user to be predicted according to the relevance through a behavior prediction model.

In one embodiment, the above apparatus further comprises:

the category selection module is used for determining the category of the target product according to the behavior prediction result;

the product determining module is used for determining a target product corresponding to the target product category in preset category products;

and the product pushing module is used for pushing the target product to the user to be predicted.

In one embodiment, the above apparatus further comprises:

the user classification module is used for classifying the users to be predicted according to the user behavior theme data to obtain user categories corresponding to the users to be predicted;

the target user determining module is used for determining a target user from the users to be predicted according to the user category and a preset screening condition;

the theme data extraction module is used for extracting target theme data corresponding to a target user from the user behavior theme data;

the behavior prediction module 808 is further configured to input the target theme data into a pre-trained behavior prediction model to obtain a behavior prediction result corresponding to the target user.

The modules in the user behavior prediction device may be implemented in whole or in part by software, hardware, and a combination thereof. The modules can be embedded in a hardware form or independent from a processor in the computer device, and can also be stored in a memory in the computer device in a software form, so that the processor can call and execute operations corresponding to the modules.

In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as shown in fig. 9. The computer device includes a processor, a memory, an Input/Output interface (I/O for short), and a communication interface. The processor, the memory and the input/output interface are connected through a system bus, and the communication interface is connected to the system bus through the input/output interface. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of an operating system and computer programs in the non-volatile storage medium. The database of the computer device is used for storing data such as report query behavior data and service behavior data. The input/output interface of the computer device is used for exchanging information between the processor and an external device. The communication interface of the computer device is used for connecting and communicating with an external terminal through a network. The computer program is executed by a processor to implement a method of user behavior prediction.

Those skilled in the art will appreciate that the architecture shown in fig. 9 is merely a block diagram of some of the structures associated with the disclosed aspects and is not intended to limit the computing devices to which the disclosed aspects apply, as particular computing devices may include more or less components than those shown, or may combine certain components, or have a different arrangement of components.

In one embodiment, a computer device is provided, comprising a memory and a processor, the memory having stored therein a computer program, the processor implementing the steps of the above-described method embodiments when executing the computer program.

In an embodiment, a computer-readable storage medium is provided, on which a computer program is stored, which computer program, when being executed by a processor, carries out the steps of the above-mentioned method embodiments.

In an embodiment, a computer program product is provided, comprising a computer program which, when being executed by a processor, carries out the steps of the above-mentioned method embodiments.

It should be noted that, the user information (including but not limited to user equipment information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) referred to in the present application are information and data authorized by the user or sufficiently authorized by each party, and the collection, use and processing of the related data need to comply with the relevant laws and regulations and standards of the relevant country and region.

It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by hardware instructions of a computer program, which can be stored in a non-volatile computer-readable storage medium, and when executed, can include the processes of the embodiments of the methods described above. Any reference to memory, database, or other medium used in the embodiments provided herein may include at least one of non-volatile and volatile memory. The nonvolatile Memory may include Read-Only Memory (ROM), magnetic tape, floppy disk, flash Memory, optical Memory, high-density embedded nonvolatile Memory, resistive Random Access Memory (ReRAM), magnetic Random Access Memory (MRAM), ferroelectric Random Access Memory (FRAM), phase Change Memory (PCM), graphene Memory, and the like. Volatile Memory can include Random Access Memory (RAM), external cache Memory, and the like. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM), among others. The databases referred to in various embodiments provided herein may include at least one of relational and non-relational databases. The non-relational database may include, but is not limited to, a block chain based distributed database, and the like. The processors referred to in the various embodiments provided herein may be, without limitation, general purpose processors, central processing units, graphics processors, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, or the like.

All possible combinations of the technical features in the above embodiments may not be described for the sake of brevity, but should be considered as being within the scope of the present disclosure as long as there is no contradiction between the combinations of the technical features.

The above-mentioned embodiments only express several embodiments of the present application, and the description thereof is more specific and detailed, but not construed as limiting the scope of the present application. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the concept of the present application, which falls within the scope of protection of the present application. Therefore, the protection scope of the present application shall be subject to the appended claims.

Claims

1. A method for predicting user behavior, the method comprising:

acquiring report query behavior data of a user to be predicted;

acquiring service behavior data of a service user;

2. The method of claim 1, wherein obtaining report query behavior data of the user to be predicted comprises:

acquiring report query behavior data of a user to be predicted from a plurality of self-service channels; the plurality of self-service channels comprise preset application programs and self-service equipment.

3. The method of claim 1, wherein the aggregating the report query behavior data and the business behavior data to obtain user behavior topic data comprises:

4. The method according to claim 1, wherein the inputting the user behavior topic data into a pre-trained behavior prediction model to obtain the behavior prediction result corresponding to the user to be predicted comprises:

5. The method of claim 1, further comprising:

and pushing the target product to the user to be predicted.

6. The method of claim 1, wherein prior to said inputting said user behavior topic data into a pre-trained behavior prediction model, said method further comprises:

7. A user behavior prediction apparatus, the apparatus comprising:

the query data acquisition module is used for acquiring report query behavior data of the user to be predicted;

the data aggregation module is used for aggregating the report query behavior data and the service behavior data to obtain user behavior topic data;

8. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor, when executing the computer program, implements the steps of the method of any of claims 1 to 6.

9. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the method of any one of claims 1 to 6.

10. A computer program product comprising a computer program, characterized in that the computer program, when being executed by a processor, carries out the steps of the method of any one of claims 1 to 6.