CN114372199A

CN114372199A - Business pushing method and device, storage medium and electronic equipment

Info

Publication number: CN114372199A
Application number: CN202210023648.5A
Authority: CN
Inventors: 梁振斌
Original assignee: Industrial and Commercial Bank of China Ltd ICBC
Current assignee: Industrial and Commercial Bank of China Ltd ICBC
Priority date: 2022-01-10
Filing date: 2022-01-10
Publication date: 2022-04-19

Abstract

The application discloses a service pushing method and device, a storage medium and electronic equipment, and relates to the field of artificial intelligence. The method comprises the following steps: acquiring a target user portrait, wherein the target user portrait is used for representing characteristic information of a target user; constructing a target AI virtual robot according to the target user portrait, and putting the target AI virtual robot into a virtual scene; predicting the demand of a target user in a real scene according to the demand of a target AI virtual robot in a virtual scene; determining a first service and a target service associated with the first service according to the requirement of a target user in a real scene, wherein the first service is a service required by a predicted target user; and pushing the target service for the target user according to a preset pushing mode. By the method and the device, the problems that the accuracy of service pushing is low and user experience is influenced in the related technology are solved.

Description

Business pushing method and device, storage medium and electronic equipment

Technical Field

The application relates to the field of artificial intelligence, in particular to a service pushing method and device, a storage medium and electronic equipment.

Background

The service recommendation is an important link of the financial institution operation, and the good pushing system can not only enhance the affinity of the customer to the financial institution, but also enable the user to enjoy good service experience. Moreover, the number and quality of customers are critical to the operation of the financial institution. Therefore, the system provides good service for the users, enables the users to be satisfied and happy, and is a hard reason for sustainable development of financial institutions.

In the related art, each large financial institution usually recommends services for users in the following recommendation manners: adopting manual hall pushing or salesman home-going pushing and telephone service pushing; based on a traditional direct marketing mode, recommending services approved by a product manager to a user on line by a stock of brains; pushing popular services obtained based on current big data analysis to users in modes of App, WeChat public numbers, web pages and the like; according to a conventional personalized pushing algorithm, services conforming to time elements are pushed for users on special dates such as holidays, birthdays, anniversaries and the like. However, the above push methods are generally based on experience judgment and market analysis of the service manager, and a push method known by the service manager is selected to push the service. Therefore, the push method in the related art does not fully consider the personal experience of the user to a great extent at present, so that the service push is a burden and a harassment for the user, and simultaneously, the resource waste and the labor consumption are caused. Moreover, more and more users start to actively shield the business from the financial institution under the bombing of the above various pushing manners, which inevitably causes the users to feel bored, and also causes the user loss and the resource waste.

Aiming at the problems that the accuracy of service push in the related technology is low and the user experience is influenced, an effective solution is not provided at present.

Disclosure of Invention

The present application mainly aims to provide a service pushing method and apparatus, a storage medium, and an electronic device, so as to solve the problem that the accuracy of service pushing is low and user experience is affected in the related art.

In order to achieve the above object, according to an aspect of the present application, a service push method is provided. The method comprises the following steps: acquiring a target user portrait, wherein the target user portrait is used for representing characteristic information of a target user; constructing a target AI virtual robot according to the target user portrait, and putting the target AI virtual robot into a virtual scene; predicting the demand of the target user in a real scene according to the demand of the target AI virtual robot in the virtual scene; determining a first service and a target service associated with the first service according to the requirement of the target user in a real scene, wherein the first service is a service required by the target user in prediction; and pushing the target service for the target user according to a preset pushing mode.

Further, the target service acquisition mode includes one of the following: judging whether the first service is a service in a service library, wherein the service library is a set for storing a plurality of services; if the first service is the service in the service library, extracting the service related to the first service from the service library, and obtaining the target service from the service related to the first service; if the first service is not the service in the service library, generating a second service, and making a decision on the second service, wherein the second service is a service related to the first service; determining whether to use the second service as the target service in accordance with a decision made for the second service.

Further, obtaining the target user representation includes: acquiring target information of a target user, wherein the target information is used for representing the attribute of the target user; analyzing the target information according to a deep learning network structure to obtain an analysis result; according to the analysis result, tagging the characteristic information of the target user to obtain first target information, wherein the first target information is information obtained by tagging the characteristic information of the target user; constructing a user model of the target user according to the first target information; and obtaining the target user portrait according to the user model of the target user and a three-dimensional modeling technology.

Further, after determining a first service and a target service associated with the first service, the method further comprises: voting the target service by a plurality of AI virtual robots by adopting a decision tree algorithm to obtain a voting result; and determining the demand degree of the target user for the target service according to the voting result.

Further, after determining the demand degree of the target user for the target service according to the voting result, the method further includes: determining cost data corresponding to the target service; and calculating to obtain a first numerical value according to the cost data corresponding to the target service and the profit brought by the target AI virtual robot for selecting the target service, wherein the first numerical value is used for expressing the estimated profit brought by the target service.

Further, after the target service is pushed to the target user according to a preset pushing manner, the method further includes: acquiring the service condition of the target user to the target service; if the target user does not use the target service, recalculating the first numerical value until the target user uses the target service; and determining the pushing effect of the target service according to the value recalculated to the first value.

Further, after determining the push effect of the target service, the method further includes: analyzing the target service according to the pushing effect of the target service; if the fact that the first user has the requirement on the target service is analyzed, and the profit of the target service meets the preset requirement, pushing the target service for the first user according to a target pushing mode, wherein the first user is other users except the target user; and if the target user fails to use the target service is detected, determining the reason why the target user fails to use the target service, and optimizing the target service according to the reason.

In order to achieve the above object, according to another aspect of the present application, a service push apparatus is provided. The device includes: the system comprises a first acquisition unit, a second acquisition unit and a third acquisition unit, wherein the first acquisition unit is used for acquiring a target user portrait, and the target user portrait is used for representing characteristic information of a target user; the first construction unit is used for constructing a target AI virtual robot according to the target user portrait and putting the target AI virtual robot into a virtual scene; the first prediction unit is used for predicting the demand of the target user in a real scene according to the demand of the target AI virtual robot in the virtual scene; a first determining unit, configured to determine, according to a requirement of the target user in a real scene, a first service and a target service associated with the first service, where the first service is a service required by the predicted target user; and the first pushing unit is used for pushing the target service for the target user according to a preset pushing mode.

Further, the target service acquisition mode includes one of the following: the first judging unit is used for judging whether the first service is a service in a service library, wherein the service library is a set for storing a plurality of services; a first processing unit, configured to, if the first service is a service in the service library, extract a service related to the first service from the service library, and obtain the target service from the service related to the first service; a second processing unit, configured to generate a second service and make a decision for the second service if the first service is not a service in the service library, where the second service is a service related to the first service; a second determining unit, configured to determine whether to use the second service as the target service according to a decision made on the second service.

Further, the first acquisition unit includes: the system comprises a first acquisition module, a second acquisition module and a third acquisition module, wherein the first acquisition module is used for acquiring target information of a target user, and the target information is used for representing the attribute of the target user; the first analysis module is used for analyzing the target information according to a deep learning network structure to obtain an analysis result; the first processing module is used for tagging the characteristic information of the target user according to the analysis result to obtain first target information, wherein the first target information is information obtained by tagging the characteristic information of the target user; the first construction module is used for constructing a user model of the target user according to the first target information; and the first obtaining module is used for obtaining the target user portrait according to the user model of the target user and a three-dimensional modeling technology.

Further, the apparatus further comprises: the system comprises a first voting unit, a second voting unit and a third voting unit, wherein the first voting unit is used for voting a target service by a plurality of AI virtual robots by adopting a decision tree algorithm after determining the first service and the target service related to the first service to obtain a voting result; and the third determining unit is used for determining the demand degree of the target user for the target service according to the voting result.

Further, the apparatus further comprises: a fourth determining unit, configured to determine, after determining, according to the voting result, a degree of demand of the target user for the target service, cost data corresponding to the target service; and the first calculating unit is used for calculating to obtain a first numerical value according to the cost data corresponding to the target service and the profit brought by the target AI virtual robot for selecting the target service, wherein the first numerical value is used for representing the pre-estimated profit brought by the target service.

Further, the apparatus further comprises: a second obtaining unit, configured to obtain a service condition of the target service for the target user after the target service is pushed to the target user according to a preset pushing manner; a second calculating unit, configured to recalculate the first value until the target user uses the target service if the target user does not use the target service; and a fifth determining unit, configured to determine a pushing effect of the target service according to the value obtained by recalculating the first value.

Further, the apparatus further comprises: the first analysis unit is used for analyzing the target service according to the pushing effect of the target service after the pushing effect of the target service is determined; a third processing unit, configured to, if it is analyzed that a first user has a demand for the target service and a benefit of the target service meets a preset requirement, push the target service to the first user according to a target push manner, where the first user is another user except the target user; and the fourth processing unit is used for determining the reason of the target service failure of the target user if the target service failure of the target user is detected, and optimizing the target service according to the reason.

In order to achieve the above object, according to another aspect of the present application, there is provided a computer-readable storage medium including a stored program, wherein the program performs the service push method of any one of the above.

In order to achieve the above object, according to another aspect of the present application, there is provided an electronic device comprising one or more processors and a memory for storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors are caused to implement the service push method of any one of the above.

Through the application, the following steps are adopted: acquiring a target user portrait, wherein the target user portrait is used for representing characteristic information of a target user; constructing a target AI virtual robot according to the target user portrait, and putting the target AI virtual robot into a virtual scene; predicting the demand of a target user in a real scene according to the demand of a target AI virtual robot in a virtual scene; determining a first service and a target service associated with the first service according to the requirement of a target user in a real scene, wherein the first service is a service required by a predicted target user; the target service is pushed to the target user according to the preset pushing mode, and the problems that the accuracy of service pushing in the related technology is low and user experience is influenced are solved. The method comprises the steps of predicting the requirements of a target user in a real scene according to the requirements of a constructed target AI virtual robot in a virtual scene, and pushing services for the target user according to the requirements of the target user and a preset pushing mode, so that the accuracy of service pushing is improved, and further the user experience is improved.

Drawings

The accompanying drawings, which are incorporated in and constitute a part of this application, illustrate embodiments of the application and, together with the description, serve to explain the application and are not intended to limit the application. In the drawings:

fig. 1 is a flowchart of a service push method provided according to an embodiment of the present application;

FIG. 2 is a personalized image of user A in an embodiment of the present application;

FIG. 3 is a personalized image of user B in an embodiment of the present application;

fig. 4 is a schematic diagram illustrating an AI virtual robot corresponding to a user a being placed in a three-dimensional virtual scene in an embodiment of the present application;

fig. 5 is a schematic diagram illustrating an AI virtual robot corresponding to a user B is placed in a three-dimensional virtual scene in an embodiment of the present application;

fig. 6 is a schematic diagram of a logical thinking model of an AI virtual robot in the embodiment of the present application;

FIG. 7 is a schematic diagram of a deep learning network model in an embodiment of the present application;

FIG. 8 is a schematic diagram of cost estimation in an embodiment of the present application;

fig. 9 is a schematic diagram of updating a service in an embodiment of the present application;

fig. 10 is a schematic diagram of a service push apparatus provided according to an embodiment of the present application;

fig. 11 is a schematic diagram of a service push system provided according to an embodiment of the present application;

fig. 12 is a schematic diagram of an electronic device provided according to an embodiment of the application.

Detailed Description

It should be noted that the embodiments and features of the embodiments in the present application may be combined with each other without conflict. The present application will be described in detail below with reference to the embodiments with reference to the attached drawings.

In order to make the technical solutions better understood by those skilled in the art, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only partial embodiments of the present application, but not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.

It should be noted that the terms "first," "second," and the like in the description and claims of this application and in the drawings described above are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It should be understood that the data so used may be interchanged under appropriate circumstances such that embodiments of the application described herein may be used. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed, but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.

It should be noted that, the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for presentation, analyzed data, etc.) referred to in the present disclosure are information and data authorized by the user or sufficiently authorized by each party.

The present invention is described below with reference to preferred implementation steps, and fig. 1 is a flowchart of a service push method provided according to an embodiment of the present application, and as shown in fig. 1, the method includes the following steps:

step S101, obtaining a target user portrait, wherein the target user portrait is used for representing feature information of a target user.

For example, a user representation conforming to the user characteristics is constructed in a three-dimensional virtual space by using a three-dimensional modeling technique according to a user model of a user, and the user representation is acquired. For example, the personalized image of the user a is shown in fig. 2, and the personalized image of the user B is shown in fig. 3.

And step S102, constructing a target AI virtual robot according to the target user portrait, and putting the target AI virtual robot into a virtual scene.

For example, according to the user portrait or the user model, the virtual is converted into an intelligent AI-Alpha person by accessing unit (interpretation and Interaction technology) technology, and the intelligent AI-Alpha person is placed in a virtual scene created according to a real scene for training, and this step is a process of constructing an intelligent AI. In addition, the scene is a simplified 3D model simulation scene created by Unity3D, and the scene model is not provided with rich and clear beautiful scenes except essential basic facilities such as houses, automobiles, shopping malls, schools and other buildings. The scene mainly provides a growing platform for a plurality of Alpha people. Meanwhile, the personality and self-consciousness of Alpha can be continuously improved through an incremental learning means. For example, a schematic diagram of placing a virtual robot corresponding to the user a into a three-dimensional virtual scene is shown in fig. 4, and a schematic diagram of placing a virtual robot corresponding to the user B into a three-dimensional virtual scene is shown in fig. 5.

And step S103, predicting the demand of the target user in the real scene according to the demand of the target AI virtual robot in the virtual scene.

For example, the real needs of the user in the real world are inferred from the actual needs of the Alpha people living in the virtual scene as described above. However, the intelligent AI can simulate some attributes of a real user to some extent, but cannot predict various complex situations in real life, and thus, the intelligent AI makes up for the deficiency. The application introduces a network structure-oriented database Neo4j, which can effectively store interpersonal relationship maps and transaction maps of Alpha people. And the attention point and the emphasis point of the user in the current state can be described by changing the connection relation and the weight in the map. In addition, as shown in fig. 6, a model diagram of the method is shown, in which a real-time mark line A, B is used to ensure real-time interaction and to simulate the inspirational sensibility of real persons, a multi-stream parallel scanning technique is adopted, and the scanning speeds and decision priorities of a plurality of parallel scanning lines are obtained by real-time incremental learning and have uncertainty. Thereby more realistically describing the emergency situations that the user may encounter in real scenes. Therefore, it can be deduced that the user A wants to buy a pair of high-end princess crystal shoes according to fig. 4, and the user B wants to buy a new house according to fig. 5, so that the services respectively required by the user A and the user B can be deduced according to the requirements of the user A and the user B.

In addition, the logic thinking of the user is a key factor for determining the preference degree of the user. In the conventional service pushing method, a service is generally formulated based on unilateral requirements of a financial institution, and the service is pushed to a user by adopting the description method in the background. In this mode, the financial institution is the provider of the service, being the master; the user is the receiver of the service and is the passive party. This leads to the common phenomenon that services provided by financial institutions are not intended by some users, and services intended by some users are not offered to the push by the financial institutions. The method and the system convert thinking, use the user as an active party, establish the logical thinking of the user through the user portrait, convert the user portrait from the characteristic description of the user into an intelligent AI capable of actively thinking, and request the service needed by the service push service system.

Step S104, determining a first service and a target service associated with the first service according to the requirement of a target user in a real scene, wherein the first service is a service required by a predicted target user;

for example, after the business service required by the user is calculated, the business related to the business required by the user is further determined, and the determined business related to the business required by the user is delivered to a business matching decision system for voting decision, and finally a business which best meets the user requirement is selected.

And step S105, pushing the target service for the target user according to a preset pushing mode.

For example, after determining the service that best meets the user's requirements, submitting the pushing process of the service to a service manager familiar with the user for review, and selecting a suitable pushing mode as a recommendation item. And the service manager selects a push mode which accords with the current practical situation in the push mode according to experience, and pushes the push mode to the user accurately. The application combines an intelligent pushing system to quantitatively push services which are urgently needed by a user in a fixed-point positioning manner. Through the favorite push mode of the user, the accurate service is provided for the user, the service mode not only can meet the actual requirements of the user, but also can reduce the waste of resources, and reduce the burden of the business selected by the user, so that the user can experience good personalized service, more clients are reserved, and more profits are brought to financial institutions. For example, after receiving the task, the business manager analyzes that the user A likes to answer the business service recommendation in a telephone mode, so that the customer service C of the delegator with beautiful voice and sweet sends the business to the user A; it is analyzed that user B prefers to receive business service recommendations in a face-to-face manner, and thus the customer service D, which is a sweetish customer service, pushes the business to user B in a face-to-face manner.

Through the steps from S101 to S105, the demand of the target user in the real scene is predicted according to the demand of the constructed target AI virtual robot in the virtual scene, and the service is pushed to the target user according to the demand of the target user and the preset pushing mode, so that the accuracy of service pushing is improved, and further the user experience is improved.

Optionally, in the service push method provided in the embodiment of the present application, the target service acquisition mode includes one of the following: judging whether the first service is a service in a service library, wherein the service library is a set for storing a plurality of services; if the first service is a service in the service library, extracting the service related to the first service from the service library, and obtaining a target service from the service related to the first service; if the first service is not a service in the service library, generating a second service, and making a decision on the second service, wherein the second service is a service related to the first service; determining whether to use the second service as a target service according to the decision made on the second service.

For example, after the business service required by the user is deduced, the business is further searched and generated. Specifically, first, the corresponding service terms are searched in the existing service library, and are corrected in an auxiliary manner. If the service is not available in the local service library, the corresponding service needs to be drawn up and submitted to wait for the decision of the manager and other related personnel, i.e. if the service is not available in the service system, a plurality of AI voting modes are adopted to decide whether the service should be provided, and if the service needs to be provided, the service is submitted to the product manager for decision; if the information is not needed to be provided, the information is stored in a database and is waited for related personnel to browse and inquire. If the corresponding service exists in the service local library, the corresponding service is extracted and submitted to a service matching decision system for voting decision. In addition, the application adopts a distributed search method, namely an elastic search, in the service search technology, wherein the method is based on a Lucene distributed search engine. Moreover, the method is different from a common serial search method, is a parallel search method based on a plurality of processors, and can greatly accelerate the data retrieval speed. In the service generation module, an open-source Python crawler technology is adopted, relevant service services are firstly searched from the Internet, then word and sentence analysis is carried out on the services, a relevance questionnaire is established, and the generation of the service is completed according to the table content. For example, the database is searched to find that the existing service business has no related business about 'crystal shoes', so that a scheme suitable for the user is compiled by searching related materials on the internet, cash consumption reserve money with the value of 20000 yuan is provided for the user, the scheme is charged into a credit card of the user in a direct charging mode, and the scheme is submitted to a business manager for judgment; finding out that the house purchasing loan service of the city where the user B is located exists in the existing service by searching the database, the pushing system selects three optimal loan schemes to send to the service manager, and recommends a pushing mode suitable for the user to be selected by the service manager.

Through the scheme, the manpower resource can be saved, the crowd can be rapidly bought, and the working efficiency is improved.

Optionally, in the service push method provided in the embodiment of the present application, acquiring a target user representation includes: acquiring target information of a target user, wherein the target information is used for representing the attribute of the target user; analyzing the target information according to the deep learning network structure to obtain an analysis result; according to the analysis result, tagging the characteristic information of the target user to obtain first target information, wherein the first target information is information obtained by tagging the characteristic information of the target user; constructing a user model of a target user according to the first target information; and obtaining the target user portrait according to the user model of the target user and the three-dimensional modeling technology.

For example, the information of the user needs to be collected, the basic personal information and the account status information of the user can be extracted according to the historical records of the counter service or the online service transacted by the user in the bank, and the basic data is stored in the relational model database to be used as the basis for describing the user. And under the permission condition of the user, the OpenCV graphic image processing technology is utilized to identify and classify the images, photos and the like of the user. And constructing the basic attribute information of the user according to the same facies and approximately having similar characters. Meanwhile, the web crawler technology is adopted to search the content and description of the user on the Internet, extract key data, construct an interpersonal relationship map and a character map, and an RSA encryption algorithm is adopted in the whole process to protect the basic information of the user from being leaked. For example, a part of the obtained user information is shown in table 1.

TABLE 1 characteristic information Table of similar users

	Feature 1	Feature 2	…	Feature 20	Label (R)
						User 1	x1	y1	z1	0
User 2	x2	y2		z2	1
						…	…	…	…	…
User 100	x100	y100	…	z100	0

The user data collected in the table above is used to package the data into three groups, one group A for training the model, one group B for testing the model, and one group C for testing the model. A. B, C, three sets of data are as follows: 2: 1, randomly extracting and adopting multiple random adjustments to detect whether the deeply learned network model is close to the optimal solution. In the training, a deep learning network structure as shown in fig. 7 is adopted, and the Loss function is a CIOU Loss function:

the IOU represents the logarithmic ratio of the attribute relevance and the attribute associativity of the users, and the larger the ratio is, the more similar the attributes among the users are, and the smaller the ratio is, the opposite is. Rho²Representing Euclidean distance between sample data, c representing difference value between prediction model and hypothesis model, b^gtRepresenting the central description point of the prediction model. To process the data more efficiently, the data is combined into Batch for training. In this way, the user's preferences, habits, personalities, ideas, etc. are tagged and stored in the database for generating the user model. According to the user model, a user portrait conforming to the user characteristics is constructed in a three-dimensional virtual space by using a three-dimensional modeling technology. The real and virtual one-to-one mapping is achieved through the model establishment of the user, and the portrayal of the user is realized.

Additionally, the user representation is a database describing the user's basic features. In the traditional mode, the description of the user stays basically on the identity information of the user and in the impression of the hall manager who receives the customer. Although the basic information can describe the user characteristics to a certain extent, the basic information can also be used as the fundamental dependence of service recommendation. However, if the service recommendation is successful or not, it is completely unexpected. The deep learning algorithm is adopted, and the actions of the user are all used as data sources for characterizing the characteristics of the user. Such as the consumption habit, financial management habit, transfer habit, trip habit, dressing habit, weather preference, eating habit, body building and beauty treatment of the user, the preference degree of the user to service pushing and the like, so as to establish a complete user portrait.

By the scheme, the data of the user can be effectively prevented from being stored from a single angle, so that the purposes of knowing the user in multiple dimensions and deducing the requirement of the user are achieved. Meanwhile, the effects of knowing the user from the reality and analyzing the user from the virtual scene can be achieved. In addition, the constructed awareness of the intelligent AI robot can lead the characters in the scene to carry out autonomous communication and communication simulation, so that the requirements of real users on services can be simulated. The personality characteristics and the preference of the user can be calculated through the omnibearing user portrait, and the firmest and most comprehensive data dependence is provided for the service recommendation system.

Optionally, in the service pushing method provided in the embodiment of the present application, after determining the first service and the target service associated with the first service, the method further includes: voting is carried out on the target service through a plurality of AI virtual robots by adopting a decision tree algorithm to obtain a voting result; and determining the demand degree of the target user for the target service according to the voting result.

For example, when the degree of the service demand of the user is judged, the finally selected service which best meets the requirement of the user, namely the finally determined service which is ready to be pushed to the user, is tested by using the Alpha person generated in the above steps. And finally, a plurality of Alpha people adopt a decision tree algorithm to perform voting selection through a plurality of batches of random selection tests, and a maximum value is selected from all optimal values. The algorithm has the following calculation formula:

in addition, the phenomenon of overcasting and abandoning in the voting process is avoided. (the over-casting means that all Alpha people irreconcilably cast yes or no votes due to certain factors in the voting process, and the over-casting means that the Alphas cannot vote under the current conditions to cause the illusion of total vote casting.) the voting process adopts a pre-pruning algorithm and a post-pruning means to ensure the voting effectiveness. For example, the proposed scheme corresponding to the user A is put into a virtualization scene, and the virtual robot corresponding to the user A immediately adopts the scheme and purchases crystal shoes with the value of 19888 yuan; the scheme corresponding to the user B is put into a virtualization scene, the virtual robot corresponding to the user B is found to be hesitant to adopt the scheme for a long time, a new house is purchased in a loan mode, and the day of eating soil is passed.

Through the scheme, the subjectivity of manual judgment is effectively avoided, and the user requirements can be met practically.

Optionally, in the service pushing method provided in this embodiment of the present application, after determining, according to the voting result, a degree of demand of the target user for the target service, the method further includes: determining cost data corresponding to the target service; and calculating to obtain a first numerical value according to the cost data corresponding to the target service and the profit brought by the target AI virtual robot for selecting the target service, wherein the first numerical value is used for expressing the estimated profit brought by the target service.

For example, after determining the degree of the user's demand for the service, cost estimation and profit estimation are required. I.e. the corresponding cost price P needs to be calculated. And the cost estimation flow chart is shown in fig. 8. And after the corresponding business service is established, the service needs to be evaluated for income, the evaluation loss function adopts a Sigmoid function, when the evaluation value is greater than 0.5, the positive income can be acquired with high probability, otherwise, the risk of loss is considered. And the function expression is as follows:

for example, by analyzing the service push cost and the profit brought by the virtual robot corresponding to the user a, the profit valuation ratio is calculated to be 0.8, and therefore the positive profit is determined to be the approximate rate; and calculating the profit valuation ratio to be 0.3 by analyzing the business pushing cost and the profit brought by the virtual robot corresponding to the user B, so that the profit is judged to be the negative profit with the large probability.

In conclusion, through risk assessment, unnecessary service pushing can be effectively avoided, so that the cost is saved, and more valuable services are provided for users.

Optionally, in the service pushing method provided in this embodiment of the present application, after pushing the target service to the target user according to a preset pushing manner, the method further includes: acquiring the use condition of a target user on a target service; if the target user does not use the target service, recalculating the first numerical value until the target user uses the target service; and determining the pushing effect of the target service according to the value recalculated to the first value.

For example, push effectiveness assessment: after the service is pushed to the user, the service use condition of the user needs to be tracked in real time and for a long time, and the service use condition is recorded as a report document. And further extracting key data by using the acquired observation data in the report document, and sending the key data to the step of collecting the user information for incremental learning, thereby further enriching the user attributes. And then continuing the next steps, recalculating and predicting the pushing of the service, continuously correcting the pushing by adopting the actual data, finally giving the evaluation data beta of the service pushing effect after the user uses the service, and judging the quality of the pushing effect according to the evaluation data beta. For example, it can be analyzed from the observation record report that the user a receives the push of the service and uses the service to purchase the crystal shoes of the heart instrument two days later, so that the evaluation result β is 0.93 according to the push effect, which is in line with the expectation; user B hesitates to receive the service for a long time and uses the service to prepare for purchasing a new house half a year after discussing the business volume, and the purchase fails due to policy reasons, so that the evaluation result β is 0.23 according to the pushing effect, which is in line with expectations.

In conclusion, by combining the virtual line and the real line, accurate service pushing can be fully achieved, and discomfort and conflict of users are avoided. And the AI in the virtual scene can be well guided to approach the real user continuously, so that data dependence is further provided for next service pushing.

Optionally, in the service pushing method provided in the embodiment of the present application, after determining the pushing effect of the target service, the method further includes: analyzing the target service according to the pushing effect of the target service; if the fact that the first user has the requirement on the target service is analyzed, and the income of the target service meets the preset requirement, the target service is pushed to the first user according to a target pushing mode, wherein the first user is other users except the target user; and if the target service using failure of the target user is detected, determining the reason of the target service using failure of the target user, and optimizing the target service according to the reason.

For example, after the push service is refined, it is necessary to determine the consequences of the acceptance or non-acceptance of the service by the user. And performing further feedback analysis on the service from generation to pushing and evaluation schemes, thereby completing pushing of the service. And a flow chart for updating the traffic in the system is shown in fig. 9. For example, by analyzing the case of the user a, it can be found that the user a1, the user a2, the user A3, and the like have the same requirements, and the recommendation system can adopt different pushing means and schemes according to respective user images to realize the popularization of the service; by analyzing the case of the user B, it can be found that the policy in the real scene is also required as an important factor for influencing the acceptance of the service.

By the scheme, the system for pushing the service can be updated iteratively, so that the service pushing range is expanded more accurately.

In summary, the basic information and the personalized attribute description of the user can be fully utilized by the method, the requirement of the uniqueness of the user is fully considered from the perspective of the user, and the service meeting the requirement of the user is customized according to the requirement. Therefore, the method and the device not only enable users to feel relieved, but also can fundamentally reduce the waste of cost and achieve good economic benefits. In addition, a table of the push effect comparison table, as shown in table 3, can be obtained by performing statistical comparison analysis with the conventional push method. As can be seen from the table, the accurate pushing mode is far superior to that of the traditional mode, the pushing period can be shortened, the pushing effect can be improved, and the income can be further enlarged, so that the user can feel more worry, and the business income is more stable.

Table 3 push effect comparison table

In summary, the service push method provided by the embodiment of the application obtains the target user portrait, wherein the target user portrait is used for representing the feature information of the target user; constructing a target AI virtual robot according to the target user portrait, and putting the target AI virtual robot into a virtual scene; predicting the demand of a target user in a real scene according to the demand of a target AI virtual robot in a virtual scene; determining a first service and a target service associated with the first service according to the requirement of a target user in a real scene, wherein the first service is a service required by a predicted target user; the target service is pushed to the target user according to the preset pushing mode, and the problems that the accuracy of service pushing in the related technology is low and user experience is influenced are solved. The method comprises the steps of predicting the requirements of a target user in a real scene according to the requirements of a constructed target AI virtual robot in a virtual scene, and pushing services for the target user according to the requirements of the target user and a preset pushing mode, so that the accuracy of service pushing is improved, and further the user experience is improved.

It should be noted that the steps illustrated in the flowcharts of the figures may be performed in a computer system such as a set of computer-executable instructions and that, although a logical order is illustrated in the flowcharts, in some cases, the steps illustrated or described may be performed in an order different than presented herein.

The embodiment of the present application further provides a service pushing device, and it should be noted that the service pushing device in the embodiment of the present application may be used to execute the method for pushing the service provided in the embodiment of the present application. The service push device provided by the embodiment of the present application is introduced below.

Fig. 10 is a schematic diagram of a service pushing apparatus according to an embodiment of the present application. As shown in fig. 10, the apparatus includes: a first acquisition unit 1001, a first construction unit 1002, a first prediction unit 1003, a first determination unit 1004, and a first push unit 1005.

Specifically, the first obtaining unit 1001 is configured to obtain a target user representation, where the target user representation is used to represent feature information of a target user;

a first constructing unit 1002, configured to construct a target AI virtual robot according to a target user portrait, and place the target AI virtual robot in a virtual scene;

the first prediction unit 1003 is used for predicting the demand of a target user in a real scene according to the demand of the target AI virtual robot in a virtual scene;

a first determining unit 1004, configured to determine, according to a requirement of a target user in a real scene, a first service and a target service associated with the first service, where the first service is a service required by a predicted target user;

a first pushing unit 1005, configured to push the target service to the target user according to a preset pushing manner.

To sum up, the service push apparatus provided in the embodiment of the present application obtains a target user portrait through the first obtaining unit 1001, where the target user portrait is used to represent feature information of a target user; the first construction unit 1002 constructs a target AI virtual robot according to a target user figure, and puts the target AI virtual robot into a virtual scene; the first prediction unit 1003 predicts the demand of a target user in a real scene according to the demand of the target AI virtual robot in a virtual scene; a first determining unit 1004 determines a first service and a target service associated with the first service according to a requirement of a target user in a real scene, wherein the first service is a service required by a predicted target user; the first pushing unit 1005 pushes the target service for the target user according to the preset pushing mode, so that the problems that the accuracy of service pushing is low and user experience is affected in the related technology are solved, the requirement of the target user in a real scene is predicted according to the requirement of the constructed target AI virtual robot in a virtual scene, and the service is pushed for the target user according to the requirement of the target user and the preset pushing mode, so that the accuracy of service pushing is improved, and further the user experience is improved.

Optionally, in the service push device provided in the embodiment of the present application, the obtaining manner of the target service includes one of the following: the first judging unit is used for judging whether the first service is a service in a service library, wherein the service library is a set for storing a plurality of services; the first processing unit is used for extracting the service related to the first service from the service library and obtaining the target service from the service related to the first service if the first service is the service in the service library; the second processing unit is used for generating a second service and making a decision on the second service if the first service is not a service in the service library, wherein the second service is a service related to the first service; and the second determining unit is used for determining whether the second service is used as the target service according to the decision made on the second service.

Optionally, in the service pushing apparatus provided in this embodiment of the present application, the first obtaining unit includes: the system comprises a first acquisition module, a second acquisition module and a third acquisition module, wherein the first acquisition module is used for acquiring target information of a target user, and the target information is used for representing the attribute of the target user; the first analysis module is used for analyzing the target information according to the deep learning network structure to obtain an analysis result; the first processing module is used for tagging the characteristic information of the target user according to the analysis result to obtain first target information, wherein the first target information is information obtained by tagging the characteristic information of the target user; the first construction module is used for constructing a user model of a target user according to the first target information; the first obtaining module is used for obtaining the target user portrait according to the user model of the target user and the three-dimensional modeling technology.

Optionally, in the service push device provided in this embodiment of the present application, the device further includes: the first voting unit is used for voting the target service by adopting a decision tree algorithm through a plurality of AI virtual robots after the first service and the target service associated with the first service are determined, so as to obtain a voting result; and the third determining unit is used for determining the demand degree of the target user for the target service according to the voting result.

Optionally, in the service push device provided in this embodiment of the present application, the device further includes: the fourth determining unit is used for determining the cost data corresponding to the target service after determining the demand degree of the target user for the target service according to the voting result; and the first calculating unit is used for calculating to obtain a first numerical value according to the cost data corresponding to the target service and the profit brought by the target AI virtual robot for selecting the target service, wherein the first numerical value is used for expressing the estimated profit brought by the target service.

Optionally, in the service push device provided in this embodiment of the present application, the device further includes: the second acquisition unit is used for acquiring the service condition of the target user to the target service after the target service is pushed to the target user according to a preset pushing mode; the second calculation unit is used for recalculating the first numerical value if the target user does not use the target service until the target user uses the target service; and the fifth determining unit is used for determining the pushing effect of the target service according to the value recalculated for the first value.

Optionally, in the service push device provided in this embodiment of the present application, the device further includes: the first analysis unit is used for analyzing the target service according to the pushing effect of the target service after the pushing effect of the target service is determined; the third processing unit is used for pushing the target service for the first user according to a target pushing mode if the fact that the first user has the requirement on the target service and the profit of the target service meets the preset requirement is analyzed, wherein the first user is other users except the target user; and the fourth processing unit is used for determining the reason of the target service failure of the target user and optimizing the target service according to the reason if the target service failure of the target user is detected.

The service pushing device comprises a processor and a memory, wherein the first acquiring unit 1001, the first constructing unit 1002, the first predicting unit 1003, the first determining unit 1004, the first pushing unit 1005 and the like are stored in the memory as program units, and the processor executes the program units stored in the memory to realize corresponding functions.

As shown in fig. 11, the schematic diagram of the service push system includes: the system comprises a user management module 101, a business service management module 102, a business generation and decision module 103, an intelligent business pushing module 104, a cost budget and pre-estimated income module 105, a Web control end and an APP control end 106. The user management module 101 is used for collecting user information, establishing a user model according to the collected user information, depicting a user portrait, and constructing an intelligent AI according to the user model or the user portrait, namely constructing an intelligent AI robot; the business service management module 102 is used for calculating the business required by the user; the service generation and decision module 103 is used for searching for services, generating services, judging the degree of service demand of users and estimating cost, and estimating profit; the intelligent service push module 104 is used for accurately pushing services; the cost budget and pre-estimated income module 105 is used for carrying out pushing effect evaluation; the Web control end and the APP control end 106 are used for completing service push.

In conclusion, by adopting the superiority of the deep learning algorithm, the problem that the accurate business service cannot be accurately provided for the client when the traditional business recommendation mode is used for pushing the service business is solved, and the intelligent business pushing system based on the user image is provided. The system can fully know the user attributes, simulate the thinking of the user according to the specific personality of the user, and deduce the actual requirements and favorite features of the user. And in turn, provide accurate services for the user, thereby effectively reducing the operating cost of the financial institution, saving human resources and enhancing the affinity between the user and the financial institution. In addition, the "precision business push system based on deep learning" mainly includes the following three functions: firstly, with the help of the intelligent learning ability of deep learning, the personality of a user is subjected to virtual modeling by deeply mining collected user behavior data, and then a user portrait close to the user behavior is generated. And the portrait is used for describing personal habits, preferences, styles, ideas, values and the like of the user, so that the user and the service user can be understood from multiple dimensions. Secondly, an intelligent AI capable of simulating the logic thinking of the user is constructed according to the established user virtual portrait model, and with the increase of the collected user behavior data, the AI virtual robot can grow gradually, and finally gradually approaches the real behavior mode of the user through continuous incremental learning, continuous virtualization deduction and practice feedback, so that the actual business requirements of the user in the current state are carved, and the financial institution is helped to expand more business services meeting the requirements of the user. And thirdly, intelligently pushing the business service which accords with the psychological expectation of the user according to the real requirement of the user so as to achieve the aim of accurately serving the client. In addition, according to the user portrait established above, the service requirement unique to each user can be known conveniently. And the pushing system judges whether to provide corresponding service pushing for the user according to a decision tree judging algorithm. Therefore, through the pushing mode, the conflict psychology of the user can be effectively avoided, the pushing cost is saved, and the expected income is achieved.

The processor comprises a kernel, and the kernel calls the corresponding program unit from the memory. The kernel can be set to be one or more than one, and the accuracy of service pushing is improved by adjusting kernel parameters.

The memory may include volatile memory in a computer readable medium, Random Access Memory (RAM) and/or nonvolatile memory such as Read Only Memory (ROM) or flash memory (flash RAM), and the memory includes at least one memory chip.

An embodiment of the present invention provides a computer-readable storage medium, on which a program is stored, where the program, when executed by a processor, implements the service push method.

The embodiment of the invention provides a processor, which is used for running a program, wherein the service pushing method is executed when the program runs.

As shown in fig. 12, an embodiment of the present invention provides an electronic device, where the device includes a processor, a memory, and a program stored in the memory and executable on the processor, and the processor executes the program to implement the following steps: acquiring a target user portrait, wherein the target user portrait is used for representing characteristic information of a target user; constructing a target AI virtual robot according to the target user portrait, and putting the target AI virtual robot into a virtual scene; predicting the demand of the target user in a real scene according to the demand of the target AI virtual robot in the virtual scene; determining a first service and a target service associated with the first service according to the requirement of the target user in a real scene, wherein the first service is a service required by the target user in prediction; and pushing the target service for the target user according to a preset pushing mode.

The processor executes the program and further realizes the following steps: the acquisition mode of the target service comprises one of the following modes: judging whether the first service is a service in a service library, wherein the service library is a set for storing a plurality of services; if the first service is the service in the service library, extracting the service related to the first service from the service library, and obtaining the target service from the service related to the first service; if the first service is not the service in the service library, generating a second service, and making a decision on the second service, wherein the second service is a service related to the first service; determining whether to use the second service as the target service in accordance with a decision made for the second service.

The processor executes the program and further realizes the following steps: obtaining the target user representation includes: acquiring target information of a target user, wherein the target information is used for representing the attribute of the target user; analyzing the target information according to a deep learning network structure to obtain an analysis result; according to the analysis result, tagging the characteristic information of the target user to obtain first target information, wherein the first target information is information obtained by tagging the characteristic information of the target user; constructing a user model of the target user according to the first target information; and obtaining the target user portrait according to the user model of the target user and a three-dimensional modeling technology.

The processor executes the program and further realizes the following steps: after determining a first service and a target service associated with the first service, the method further comprises: voting the target service by a plurality of AI virtual robots by adopting a decision tree algorithm to obtain a voting result; and determining the demand degree of the target user for the target service according to the voting result.

The processor executes the program and further realizes the following steps: after determining the demand degree of the target user for the target service according to the voting result, the method further comprises: determining cost data corresponding to the target service; and calculating to obtain a first numerical value according to the cost data corresponding to the target service and the profit brought by the target AI virtual robot for selecting the target service, wherein the first numerical value is used for expressing the estimated profit brought by the target service.

The processor executes the program and further realizes the following steps: after the target service is pushed to the target user according to a preset pushing mode, the method further includes: acquiring the service condition of the target user to the target service; if the target user does not use the target service, recalculating the first numerical value until the target user uses the target service; and determining the pushing effect of the target service according to the value recalculated to the first value.

The processor executes the program and further realizes the following steps: after determining the push effect of the target service, the method further includes: analyzing the target service according to the pushing effect of the target service; if the fact that the first user has the requirement on the target service is analyzed, and the profit of the target service meets the preset requirement, pushing the target service for the first user according to a target pushing mode, wherein the first user is other users except the target user; and if the target user fails to use the target service is detected, determining the reason why the target user fails to use the target service, and optimizing the target service according to the reason. The device herein may be a server, a PC, a PAD, a mobile phone, etc.

The present application further provides a computer program product adapted to perform a program for initializing the following method steps when executed on a data processing device: acquiring a target user portrait, wherein the target user portrait is used for representing characteristic information of a target user; constructing a target AI virtual robot according to the target user portrait, and putting the target AI virtual robot into a virtual scene; predicting the demand of the target user in a real scene according to the demand of the target AI virtual robot in the virtual scene; determining a first service and a target service associated with the first service according to the requirement of the target user in a real scene, wherein the first service is a service required by the target user in prediction; and pushing the target service for the target user according to a preset pushing mode.

When executed on a data processing device, is further adapted to perform a procedure for initializing the following method steps: the acquisition mode of the target service comprises one of the following modes: judging whether the first service is a service in a service library, wherein the service library is a set for storing a plurality of services; if the first service is the service in the service library, extracting the service related to the first service from the service library, and obtaining the target service from the service related to the first service; if the first service is not the service in the service library, generating a second service, and making a decision on the second service, wherein the second service is a service related to the first service; determining whether to use the second service as the target service in accordance with a decision made for the second service.

When executed on a data processing device, is further adapted to perform a procedure for initializing the following method steps: obtaining the target user representation includes: acquiring target information of a target user, wherein the target information is used for representing the attribute of the target user; analyzing the target information according to a deep learning network structure to obtain an analysis result; according to the analysis result, tagging the characteristic information of the target user to obtain first target information, wherein the first target information is information obtained by tagging the characteristic information of the target user; constructing a user model of the target user according to the first target information; and obtaining the target user portrait according to the user model of the target user and a three-dimensional modeling technology.

When executed on a data processing device, is further adapted to perform a procedure for initializing the following method steps: after determining a first service and a target service associated with the first service, the method further comprises: voting the target service by a plurality of AI virtual robots by adopting a decision tree algorithm to obtain a voting result; and determining the demand degree of the target user for the target service according to the voting result.

When executed on a data processing device, is further adapted to perform a procedure for initializing the following method steps: after determining the demand degree of the target user for the target service according to the voting result, the method further comprises: determining cost data corresponding to the target service; and calculating to obtain a first numerical value according to the cost data corresponding to the target service and the profit brought by the target AI virtual robot for selecting the target service, wherein the first numerical value is used for expressing the estimated profit brought by the target service.

When executed on a data processing device, is further adapted to perform a procedure for initializing the following method steps: after the target service is pushed to the target user according to a preset pushing mode, the method further includes: acquiring the service condition of the target user to the target service; if the target user does not use the target service, recalculating the first numerical value until the target user uses the target service; and determining the pushing effect of the target service according to the value recalculated to the first value.

When executed on a data processing device, is further adapted to perform a procedure for initializing the following method steps: after determining the push effect of the target service, the method further includes: analyzing the target service according to the pushing effect of the target service; if the fact that the first user has the requirement on the target service is analyzed, and the profit of the target service meets the preset requirement, pushing the target service for the first user according to a target pushing mode, wherein the first user is other users except the target user; and if the target user fails to use the target service is detected, determining the reason why the target user fails to use the target service, and optimizing the target service according to the reason.

As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.

The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

In a typical configuration, a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.

The memory may include forms of volatile memory in a computer readable medium, Random Access Memory (RAM) and/or non-volatile memory, such as Read Only Memory (ROM) or flash memory (flash RAM). The memory is an example of a computer-readable medium.

Computer-readable media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), Static Random Access Memory (SRAM), Dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), Read Only Memory (ROM), Electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), Digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information that can be accessed by a computing device. As defined herein, a computer readable medium does not include a transitory computer readable medium such as a modulated data signal and a carrier wave.

It should also be noted that the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other identical elements in the process, method, article, or apparatus that comprises the element.

The above are merely examples of the present application and are not intended to limit the present application. Various modifications and changes may occur to those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the claims of the present application.

Claims

1. A service pushing method is characterized by comprising the following steps:

acquiring a target user portrait, wherein the target user portrait is used for representing characteristic information of a target user;

constructing a target AI virtual robot according to the target user portrait, and putting the target AI virtual robot into a virtual scene;

predicting the demand of the target user in a real scene according to the demand of the target AI virtual robot in the virtual scene;

determining a first service and a target service associated with the first service according to the requirement of the target user in a real scene, wherein the first service is a service required by the target user in prediction;

and pushing the target service for the target user according to a preset pushing mode.

2. The method of claim 1, wherein the target service is obtained in a manner that includes one of:

judging whether the first service is a service in a service library, wherein the service library is a set for storing a plurality of services;

if the first service is the service in the service library, extracting the service related to the first service from the service library, and obtaining the target service from the service related to the first service;

if the first service is not the service in the service library, generating a second service, and making a decision on the second service, wherein the second service is a service related to the first service;

determining whether to use the second service as the target service in accordance with a decision made for the second service.

3. The method of claim 1, wherein obtaining a target user representation comprises:

acquiring target information of a target user, wherein the target information is used for representing the attribute of the target user;

analyzing the target information according to a deep learning network structure to obtain an analysis result;

according to the analysis result, tagging the characteristic information of the target user to obtain first target information, wherein the first target information is information obtained by tagging the characteristic information of the target user;

constructing a user model of the target user according to the first target information;

and obtaining the target user portrait according to the user model of the target user and a three-dimensional modeling technology.

4. The method of claim 1, wherein after determining a first service and a target service associated with the first service, the method further comprises:

voting the target service by a plurality of AI virtual robots by adopting a decision tree algorithm to obtain a voting result;

and determining the demand degree of the target user for the target service according to the voting result.

5. The method of claim 4, wherein after determining the demand level of the target service for the target user according to the voting result, the method further comprises:

determining cost data corresponding to the target service;

and calculating to obtain a first numerical value according to the cost data corresponding to the target service and the profit brought by the target AI virtual robot for selecting the target service, wherein the first numerical value is used for expressing the estimated profit brought by the target service.

6. The method of claim 5, wherein after pushing the target service for the target user according to a preset pushing manner, the method further comprises:

acquiring the service condition of the target user to the target service;

if the target user does not use the target service, recalculating the first numerical value until the target user uses the target service;

and determining the pushing effect of the target service according to the value recalculated to the first value.

7. The method of claim 6, wherein after determining the push effect of the target service, the method further comprises:

analyzing the target service according to the pushing effect of the target service;

if the fact that the first user has the requirement on the target service is analyzed, and the profit of the target service meets the preset requirement, pushing the target service for the first user according to a target pushing mode, wherein the first user is other users except the target user;

and if the target user fails to use the target service is detected, determining the reason why the target user fails to use the target service, and optimizing the target service according to the reason.

8. A service push apparatus, comprising:

the system comprises a first acquisition unit, a second acquisition unit and a third acquisition unit, wherein the first acquisition unit is used for acquiring a target user portrait, and the target user portrait is used for representing characteristic information of a target user;

the first construction unit is used for constructing a target AI virtual robot according to the target user portrait and putting the target AI virtual robot into a virtual scene;

the first prediction unit is used for predicting the demand of the target user in a real scene according to the demand of the target AI virtual robot in the virtual scene;

a first determining unit, configured to determine, according to a requirement of the target user in a real scene, a first service and a target service associated with the first service, where the first service is a service required by the predicted target user;

and the first pushing unit is used for pushing the target service for the target user according to a preset pushing mode.

9. A computer-readable storage medium, characterized in that the storage medium comprises a stored program, wherein the program executes the service push method according to any one of claims 1 to 7.

10. An electronic device comprising one or more processors and memory storing one or more programs, wherein the one or more programs, when executed by the one or more processors, cause the one or more processors to implement the traffic push method of any of claims 1-7.