Financial product recommendation method and system based on artificial intelligence, and electronic equipment

Abstract

The invention discloses a financial product recommendation method and system based on artificial intelligence and electronic equipment, and relates to the field of financial science and technology or other related fields, and the method comprises the steps: obtaining multi-source data of a target user and other users, carrying out the preprocessing of the multi-source data, calculating the similarity of the target user and other users, and carrying out the recommendation of the target user and other users. Determining similar users and defining financial products browsed or purchased by the similar users as intention products, calculating intention degree values of target users to the intention products based on the similarity and behaviors and consumption data of the similar users to the intention products, selecting target intention products, and determining the intention degree values of the target users to the intention products based on multi-source data of the target users. And performing credit rating on the target user by adopting the credit evaluation model, generating product recommendation information based on the target intention product and the credit level, and sending the recommendation information to the target user. The technical problem of low personalized degree of recommended financial products caused by low matching degree of recommendation algorithms and low accuracy of user credit evaluation results in related technologies is solved.

Description

Technical Field

[0001] The present invention relates to the field of financial technology or other related fields, and in particular, to an artificial intelligence-based financial product recommendation method and system, and electronic equipment. Background Art

[0002] With the rapid development of Internet technology, financial platforms have evolved from the initial single account management to a comprehensive financial service platform that includes transactions, investments, payments, loans, financial management, insurance, investment consulting and other services. The feature-rich financial platform provides users with more convenient and comprehensive financial services. When recommending financial products to users, financial platforms often need to consider data such as users' investment preferences, asset allocation status, and credit status, and analyze users' preference characteristics and credit characteristics based on these data, and then provide users with personalized recommendations and suggestions.

[0003] In the related technologies, the existing financial product recommendation strategies of financial institutions have certain defects. For example, the existing personalized recommendation algorithms for financial products are not intelligent enough. They usually recommend financial products to users based on simple preset rules. The recommended financial products do not match the actual needs and preferences of users, making it difficult to provide targeted recommendations to users. At the same time, the existing credit assessment model is relatively simple and only relies on certain indicators. The assessment results of user credit are not accurate enough, and there is a lack of tracking capabilities for dynamic changes in user credit, making it difficult to reflect the user's credit status in real time. Due to the low matching degree of the recommendation algorithm and the low accuracy of the user credit assessment results, the financial platform is unable to accurately analyze the user's preference characteristics and credit characteristics, which in turn affects the degree of personalization provided by the financial platform to users.

[0004] To address the above-mentioned problems, no effective solution has been proposed yet. Summary of the invention

[0005] The embodiments of the present invention provide a financial product recommendation method and system based on artificial intelligence, and an electronic device, so as to at least solve the technical problems in the related art that the matching degree of the recommendation algorithm is low, the accuracy of the result of the user credit assessment is low, and the recommended financial products are not personalized.

[0006] In order to achieve the above-mentioned purpose, according to one aspect of the present application, a financial product recommendation method based on artificial intelligence is provided, including: obtaining multi-source data of a target user and other users, and pre-processing the multi-source data, wherein the other users are customers of financial institutions other than the target user, and the multi-source data include: identity data, behavior data, consumption data, asset data and loan data; based on the multi-source data, calculating the similarity value between the target user and other users, determining similar users based on the similarity value, and defining the financial products browsed or purchased by similar users as intended products; based on the similarity value and the behavior data and consumption data of similar users on the intended products, calculating the intention degree value of the target user for the intended products, and selecting a target intended product for the target user from all intended products according to the intention degree value; based on the identity data, asset data and loan data of the target user, using a pre-built credit assessment model to credit rate the target user to obtain the credit level of the target user; based on the target intended product and credit level of the target user, generating product recommendation information, and sending the product recommendation information to the client held by the target user.

[0007] Optionally, before obtaining multi-source data of the target user and other users, it includes: performing real-name authentication on the target user and other users, the real-name authentication operations include: identity information verification and biometric verification, wherein identity information verification includes: name, identification document verification, biometric verification includes: fingerprint, face verification; establishing a connection with the associated accounts of the target user and other users, wherein the associated accounts include: financial institution accounts, e-commerce platform accounts and social platform accounts.

[0008] Optionally, the step of calculating the target user's intention level for the intended product includes: obtaining behavioral data of similar users for all intended products, and scoring the behavioral data based on preset rules; obtaining the time points at which similar users generated behavioral data, and calculating the time difference between the time point at which the behavioral data occurred and the current time point; for each intended product, calculating the similar user's preference value for the intended product based on the score and time difference of the corresponding behavioral data; and constructing an intended product matrix based on the similar user's preference values ​​for all intended products and the consumption data of the intended products.

[0009] Optionally, the step of calculating the target user's intention level for the intended product also includes: obtaining all financial products of the financial institution, calculating the target user's preference value for all financial products based on the target user's behavioral data, and calculating the target user's active value based on all preference values; calculating the target user's intention level for similar users' intended products based on the target user's active value, the similarity value between the target user and similar users, and the intended product matrix of similar users.

[0010] Optionally, the credit assessment model is constructed in the following manner: based on the historical identity data, historical asset data, historical loan data and historical credit levels of the target user and other users, a plurality of decision trees are randomly constructed using a machine learning strategy; a plurality of empirical decision trees input by an external terminal are obtained, wherein the empirical decision trees are constructed by the external terminal based on historical experience; the historical identity data, historical asset data and historical loan data of the target user and other users are respectively input into the randomly constructed decision tree and the empirical decision tree to obtain a plurality of output credit levels; based on all output credit levels and known historical credit levels, a machine learning strategy is used to determine the weights of the randomly constructed decision tree and the empirical decision tree; a credit assessment model is determined based on the randomly constructed decision tree and the corresponding weights, the empirical decision tree and the corresponding weights.

[0011] Optionally, a pre-built credit assessment model is used to perform credit rating on a target user to obtain the target user's credit level, including: inputting the target user's identity data, asset data, and loan data into multiple decision trees in the credit assessment model respectively to obtain multiple output credit levels; and determining the target user's credit level based on the weights and output credit levels of all decision trees.

[0012] Optionally, in addition to sending recommendation information to the target user, it also includes: sending account change information, subscription service notifications, event promotion information and market information to the target user based on the target user's behavior data, consumption data and target intended products.

[0013] According to another aspect of an embodiment of the present invention, there is also provided a financial product recommendation system based on artificial intelligence, including: a user authentication unit, used to perform real-name authentication on a target user and other users, and to establish connections with the target user and associated accounts of other users; a data management unit, used to manage multi-source data of the target user and other users, and to pre-process the multi-source data; an asset management unit, used to calculate the similarity value between the target user and other users, determine similar users of the target user, and calculate the target user's intention degree value for the intended product based on the similarity value and the behavioral data and consumption data of the similar users on the intended product, and determine the target intended product of the target user from the intended products according to the intention degree value; a user credit evaluation unit, used to perform credit rating on the target user according to the target user's identity data, asset data and loan data using a pre-built credit evaluation model to obtain the target user's credit level; a message push unit, used to generate product recommendation information according to the target user's target intended product and credit level, and send the product recommendation information to a client held by the target user.

[0014] Optionally, the user authentication unit includes: a real-name authentication module, used to perform real-name authentication on the target user and other users, the real-name authentication operations include: identity information verification and biometric verification, wherein identity information verification includes: name, identification document verification, biometric verification includes: fingerprint, face verification; an account connection module, used to establish a connection with the target user and other users' associated accounts, wherein the associated accounts include: financial institution accounts, e-commerce platform accounts, and social platform accounts.

[0015] Optionally, the data management unit includes: a management module for managing multi-source data of the target user and other users; and a pre-processing module for pre-processing the multi-source data.

[0016] Optionally, the asset management unit includes: a first calculation module, which calculates the similarity value between the target user and other users based on multi-source data, and determines similar users based on the similarity value; a definition module, which is used to define financial products browsed or purchased by similar users as intended products; a scoring module, which is used to obtain the behavioral data of similar users for all intended products, and score the behavioral data based on preset rules; a second calculation module, which is used to obtain the time point when the similar users generate behavioral data, and calculate the time difference between the occurrence time point of the behavioral data and the current time point; a third calculation module, for each intended product, calculates the preference value of the similar user for the intended product based on the score and time difference of the corresponding behavioral data; a first construction module, which constructs an intended product matrix based on the preference values ​​of similar users for all intended products and the consumption data of the intended products.

[0017] Optionally, the asset management unit also includes: a fourth calculation module, which is used to obtain all financial products of the financial institution, calculate the target user's preference value for all financial products based on the target user's behavioral data, and calculate the target user's active value based on all preference values; a fifth calculation module, which calculates the target user's intention level value for similar users' intended products based on the target user's active value, the similarity value between the target user and similar users, and the similar users' intended product matrix, and determines the target user's target intended product from the intended products according to the intention level value.

[0018] Optionally, the user credit assessment unit includes: a second construction module, which randomly constructs multiple decision trees based on historical identity data, historical asset data, historical loan data and historical credit levels of the target user and other users using a machine learning strategy; an acquisition module, which is used to obtain multiple experience decision trees input by an external terminal, wherein the experience decision trees are constructed by the external terminal based on historical experience; a first input module, which is used to input the historical identity data, historical asset data and historical loan data of the target user and other users into the randomly constructed decision trees and the experience decision trees respectively, to obtain multiple output credit levels; a first determination module, which determines the weights of the randomly constructed decision trees and the experience decision trees based on all output credit levels and known historical credit levels using a machine learning strategy; a second determination module, which determines the credit assessment model based on the randomly constructed decision trees and corresponding weights, the experience decision trees and corresponding weights.

[0019] Optionally, the user credit assessment unit also includes: a second input module, used to input the identity data, asset data and loan data of the target user into multiple decision trees in the credit assessment model respectively to obtain multiple output credit levels; a third determination module, which determines the credit level of the target user based on the weights and output credit levels of all decision trees.

[0020] Optionally, the message push unit includes: a first push module, which is used to generate product recommendation information based on the target user's target intended products and credit level, and send the product recommendation information to the client held by the target user; a second push module, which sends account change information, subscription service notifications, event promotion information and market information information to the target user based on the target user's behavior data, consumption data and target intended products.

[0021] According to another aspect of an embodiment of the present invention, a computer-readable storage medium is also provided, wherein the computer-readable storage medium includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute any one of the above-mentioned artificial intelligence-based financial product recommendation methods.

[0022] According to another aspect of an embodiment of the present invention, there is also provided an electronic device, comprising one or more processors and a memory, wherein the memory is used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement any one of the above-mentioned artificial intelligence-based financial product recommendation methods.

[0023] According to another aspect of an embodiment of the present invention, a computer program product is provided, including a computer program, which, when executed by a processor, implements the steps of any one of the above-mentioned artificial intelligence-based financial product recommendation methods.

[0024] In the present disclosure, multi-source data of a target user and other users are obtained, the multi-source data are pre-processed, and then based on the multi-source data, a similarity value between the target user and other users is calculated, similar users are determined based on the similarity value, and financial products browsed or purchased by similar users are defined as intended products, and then based on the similarity value and the behavioral data and consumption data of similar users on the intended products, the target user's intention level value for the intended products is calculated, and a target intended product for the target user is selected from all intended products according to the intention level value, and then based on the target user's identity data, asset data, and loan data, a pre-built credit assessment model is used to perform a credit rating on the target user to obtain the target user's credit level, and finally based on the target user's target intended product and credit level, product recommendation information is generated, and the product recommendation information is sent to a client held by the target user.

[0025] Based on the above-mentioned public content, by acquiring multi-source data and preprocessing it, data from different sources are effectively integrated, providing comprehensive data support for the recommendation of financial products and the credit evaluation of users. By calculating the similarity values ​​between users and generating recommended financial products based on the user's behavior data and consumption data, at the same time, by using a pre-built credit evaluation model and combining the user's identity data, asset data and loan data, the user's credit level is accurately evaluated. By combining intelligently generated financial products and accurate credit levels to recommend financial products to users, the user's preference characteristics and credit characteristics are better matched, the accuracy of the financial products recommended to users is improved, and the personalization of the recommended products is improved, thereby solving the technical problems in the related technology of low matching of recommendation algorithms, low accuracy of user credit evaluation results, and low personalization of recommended financial products. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0027] Figure 1 A hardware structure block diagram of a computer terminal (or mobile device) for implementing a financial product recommendation method based on artificial intelligence is shown;

[0028] Figure 2 is a flow chart of an optional artificial intelligence-based financial product recommendation method according to an embodiment of the present invention;

[0029] Figure 3 is a simplified flow chart of an optional artificial intelligence-based financial product recommendation method according to an embodiment of the present invention;

[0030] Figure 4 is a schematic diagram of an optional artificial intelligence-based financial product recommendation system according to an embodiment of the present invention;

[0031] Figure 5 It is a structural block diagram of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0032] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0033] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0034] It should be noted that the artificial intelligence-based financial product recommendation method and device in the present disclosure can be used in the field of financial technology (when financial products are recommended to target users based on artificial intelligence), and can also be used in any field other than the field of financial technology (when financial products are recommended to target users based on artificial intelligence). The present disclosure does not limit the application field of the artificial intelligence-based financial product recommendation method and device.

[0035] It should be noted that the information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) collected by this disclosure are information and data authorized by the user or fully authorized by all parties, and the collection, storage, use, processing, transmission, provision, disclosure and application of relevant data are in compliance with the relevant laws, regulations and standards of the relevant regions, necessary confidentiality measures are taken, and public order and good customs are not violated, and corresponding operation portals are provided for users to choose to authorize or refuse. For example, an interface is set up between this system and relevant users or organizations. Before obtaining relevant information, it is necessary to send an acquisition request to the aforementioned user or organization through the interface, and obtain relevant information after receiving the consent information fed back by the aforementioned user or organization.

[0036] It should be noted that in the present disclosure, when collecting and analyzing customer information, corresponding operation entrances are provided for users to choose to agree or reject the automated decision results; if the user chooses to reject, the expert decision-making process is entered.

[0037] The following embodiments of the present invention can be applied to various systems / applications / devices for recommending financial products based on artificial intelligence. The present invention obtains multi-source data and pre-processes it, effectively integrating data from different sources, providing comprehensive data support for the recommendation of financial products and the credit evaluation of users, and generating recommended financial products by calculating the similarity values ​​between users and based on the user's behavior data and consumption data. At the same time, by using a pre-built credit evaluation model and combining the user's identity data, asset data, and loan data, the user's credit level is accurately evaluated. By combining intelligently generated financial products and accurate credit levels to recommend financial products to users, the user's preference characteristics and credit characteristics are better matched, effectively improving the accuracy and personalization of financial product recommendations.

[0038] The present invention is described in detail below in conjunction with various embodiments.

[0039] Embodiment 1

[0040] According to an embodiment of the present invention, an embodiment of a financial product recommendation method based on artificial intelligence is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0041] The artificial intelligence-based financial product recommendation method embodiment provided in Embodiment 1 of the present application can be executed in a mobile terminal, a computer terminal or a similar computing device. Figure 1The hardware structure block diagram of a computer terminal (or mobile device) for implementing a financial product recommendation method based on artificial intelligence is shown. Figure 1 As shown, the computer terminal 10 (or mobile device) may include one or more ( Figure 1 The computer system 102 includes a processor 102 (the processor 102 may include but is not limited to a microprocessor MCU (Microcontroller Unit) or a programmable logic device FPGA (Field Programmable Gate Array)), a memory 104 for storing data, and a transmission device 106 for communication functions. In addition, it may also include: a display, an input / output interface (I / O interface), a Universal Serial Bus (USB) port (which may be included as one of the ports of the BUS bus), a network interface, a power supply and / or a camera. A person skilled in the art can understand that Figure 1 The structure shown is only for illustration and does not limit the structure of the above electronic device. Figure 1 More or fewer components as shown, or with Figure 1 Different configurations are shown.

[0042] It should be noted that the one or more processors 102 and / or other data processing circuits described above may generally be referred to herein as "data processing circuits". The data processing circuits may be embodied in whole or in part as software, hardware, firmware, or any other combination thereof. In addition, the data processing circuit may be a single independent processing module, or may be incorporated in whole or in part into any of the other components in the computer terminal 10 (or mobile device). As described in the embodiments of the present application, the data processing circuit acts as a processor control (e.g., selection of a variable resistor terminal path connected to an interface).

[0043] The memory 104 can be used to store software programs and modules of application software, such as program instructions / data storage devices corresponding to the artificial intelligence-based financial product recommendation method in the embodiment of the present application. The processor 102 executes various functional applications and data processing by running the software programs and modules stored in the memory 104, that is, realizing the above-mentioned artificial intelligence-based financial product recommendation method. The memory 104 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 104 may further include a memory remotely arranged relative to the processor 102, and these remote memories may be connected to the computer terminal 10 via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0044] The transmission device 106 is used to receive or send data via a network. The specific example of the above network may include a wireless network provided by a communication provider of the computer terminal 10. In one example, the transmission device 106 includes a network adapter (Network Interface Controller, NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 106 can be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0045] The display may be, for example, a touch screen liquid crystal display (LCD), which enables a user to interact with a user interface of the computer terminal 10 (or mobile device).

[0046] Under the above operating environment, this application provides Figure 2 The artificial intelligence-based financial product recommendation method shown. Figure 2 is a flowchart of an optional artificial intelligence-based financial product recommendation method according to an embodiment of the present invention, such as Figure 2 As shown, the method comprises the following steps:

[0047] Step S201, obtaining multi-source data of the target user and other users, and pre-processing the multi-source data, wherein the other users are customers of financial institutions other than the target user, and the multi-source data includes: identity data, behavior data, consumption data, asset data and loan data.

[0048] In an embodiment of the present invention, multi-source data of a target user and other users is first obtained, where the "target user" refers to a specific individual for whom the embodiment of the present invention is providing personalized financial product recommendation services, and the "other users" cover all or part of the customers of the financial institution except the target user.

[0049] The scope of multi-source data is wide, and it includes user identity data, behavior data, consumption data, asset data and loan data. Among them, identity data includes: age, gender, birthday, work unit, industry, education, residential address, family member relationship, social relationship, income level; behavior data includes: social behavior data, browsing or purchasing behavior data generated when the target user and other users use associated accounts, and current location data when the target user and other users use associated accounts; consumption data includes: business travel data, catering consumption data, entertainment consumption data, life consumption data, online shopping data; wealth data includes: real estate information, vehicle information, credit data, family income information, historical credit rating, insurance information; loan data includes: mortgage information, vehicle loan information, operating loan information, successful borrowing ratio, successful repayment ratio, and number of overdue loans.

[0050] After obtaining these multi-source data, they are preprocessed. The preprocessing process includes data cleaning, data conversion, and normalization. The data cleaning process removes outliers and duplicate data in the multi-source data and fills in missing values. The data conversion process converts non-numeric data into numeric data. Normalization ensures that multi-source data of the same type are on the same scale. By preprocessing multi-source data, the quality of multi-source data is improved, providing a good data foundation for the subsequent generation of recommended financial products and user credit assessment.

[0051] Step S202, based on the multi-source data, calculate the similarity value between the target user and other users, determine similar users based on the similarity value, and define the financial products browsed or purchased by the similar users as intended products.

[0052] After obtaining the multi-source data and preprocessing it, the similarity value between the target user and other users is calculated. The similarity value is a quantitative index obtained by comparing the similarity between the target user and other users on the multi-source data. The embodiment of the present invention uses a pre-built multi-layer perceptron model to calculate the similarity value between the target user and other users. The multi-layer perceptron model consists of an input layer, a hidden layer, and an output layer. The multi-source data of the target user and other users are input into the multi-layer perceptron to generate the final similarity value, and the other users with the largest similarity value are determined as similar users of the target user, that is, a class of users who have similar financial management concepts and behavioral habits to the target user.

[0053] At the same time, the embodiment of the present invention determines the financial products browsed or purchased by similar users as intended products. These intended products serve as preliminary intended products for the target user, that is, the financial products browsed or purchased by users similar to the target user are determined as preliminary intended products for the target user. Subsequently, these preliminary intended products will be further narrowed down, and finally the target intended products will be determined and recommended to the target user.

[0054] Step S203, based on the similarity value and the behavior data and consumption data of similar users on the intended product, calculate the target user's intention level value for the intended product, and select the target intended product for the target user from all intended products according to the intention level value.

[0055] After determining the similar users and intended products of the target user, the embodiment of the present invention further uses the behavior data and consumption data of the similar users on the intended products to calculate the target user's intention level value for the intended products, and then selects the target intended product for the target user from all intended products according to the intention level value. This process aims to quantify the target user's interest in different intended products, and then can filter out the financial products that the target user is most interested in from the intended products.

[0056] Step S204, based on the identity data, asset data and loan data of the target user, a pre-built credit assessment model is used to perform a credit rating on the target user to obtain the credit level of the target user.

[0057] In addition to screening out target intended products for target users, the embodiments of the present invention also perform credit rating on target users. Based on the identity data, asset data, and loan data of the target users, the embodiments of the present invention use a pre-built credit assessment model to perform credit rating on target users, thereby obtaining the credit level of the target users. The purpose of credit rating on target users is to accurately assess the credit status of target users, and then recommend more suitable financial products to target users. For example, for target users with good credit status, high-risk and high-yield financial products can be recommended to them, while for target users with poor credit status, savings-type financial products can be recommended to them.

[0058] The credit assessment model provided by the embodiment of the present invention is constructed and trained based on historical data. In actual use scenarios, the target user's identity data, asset data, and loan data are input into the credit assessment model to obtain the target user's credit rating. It should be noted that the embodiment of the present invention divides the user's credit level into the following categories: customers with the highest default risk, customers with a relatively high default risk, target customers, high-quality customers, customers with a relatively high potential risk, and customers with the highest potential risk.

[0059] Step S205: Generate product recommendation information based on the target intended product and credit level of the target user, and send the product recommendation information to the client held by the target user.

[0060] Finally, the embodiment of the present invention generates product recommendation information based on the target user's target intended product and credit level, and sends the product recommendation information to the client held by the target user. This process combines the target user's financial management preferences, credit status and behavior patterns to generate highly customized product recommendation information, and accurately sends these product recommendation information to the target user through the target user's client at the financial institution level, such as mobile phone applications, emails or SMS services, and then recommends financial products that match the target user's asset status, investment preferences and credit status. Through personalized product recommendations, financial institutions can provide services that better meet user needs, thereby improving the user's capital retention rate.

[0061] Optionally, before obtaining multi-source data of the target user and other users, it includes: performing real-name authentication on the target user and other users, the real-name authentication operations include: identity information verification and biometric verification, wherein identity information verification includes: name, identification document verification, biometric verification includes: fingerprint, face verification; establishing a connection with the associated accounts of the target user and other users, wherein the associated accounts include: financial institution accounts, e-commerce platform accounts and social platform accounts.

[0062] In the embodiment of the present invention, in order to ensure the data security and user privacy protection on the financial platform, the embodiment of the present invention performs real-name authentication on the target user and other users before obtaining multi-source data of the target user and other users. The operation of real-name authentication includes identity information verification and biometric verification, wherein identity information verification needs to verify whether the user's name, age, identity document and other identity information are correct, and biometric verification needs to verify whether the user's fingerprint, face and other features match the identity information.

[0063] At the same time, in order to obtain multi-source data of users, the embodiment of the present invention establishes a connection with the user's associated account before obtaining the data. The associated account refers to the account that the user has in other financial institutions, mainstream e-commerce platforms, social platforms, etc. Financial institutions perform real-name authentication on users and establish connections with the user's associated accounts to obtain comprehensive identity data, behavior data, consumption data, asset data, and loan data of users, thereby providing users with more accurate financial services.

[0064] Optionally, the step of calculating the target user's intention level for the intended product includes: obtaining behavioral data of similar users for all intended products, and scoring the behavioral data based on preset rules; obtaining the time points at which similar users generated behavioral data, and calculating the time difference between the time point at which the behavioral data occurred and the current time point; for each intended product, calculating the similar user's preference value for the intended product based on the score and time difference of the corresponding behavioral data; and constructing an intended product matrix based on the similar user's preference values ​​for all intended products and the consumption data of the intended products.

[0065] In step S203, when calculating the target user's intention level for the intended product based on the similarity value and the behavior data and consumption data of similar users for the intended product, the behavior data is first scored based on the preset rules. For example, if the similar user has only browsed the intended product, the behavior data score is 0.1 points; if the similar user has liked the intended product, the behavior data score is 0.5 points; if the similar user has purchased the intended product, the behavior data score is 1 point. At the same time, the time point when the similar user generates the behavior data is obtained, and the time difference between the occurrence time point of the behavior data and the current time point is calculated. For example, if the time point when the similar user browses, likes or purchases the intended product is 9:50 on May 5, and the current time point is 9:51 on May 5, the time difference is 60 seconds.

[0066] For each intended product, after obtaining the behavioral data score and time difference of the intended product, calculate the preference value p of similar users for the intended product ij , the calculation formula is:

[0067]

[0068] In the above formula, M ij represents the preference value of similar user i for intended product j, ε represents the behavioral data score of similar user i for intended product j, α is the preset time decay rate, t is the time difference of the behavioral data of similar user i for intended product j, Q is the set of all financial products, and M iq Represents the set of all intended products of similar user i.

[0069] Based on the above formula, the preference values ​​of similar users for all intended products are calculated, and then the intended product matrix is ​​constructed by combining the consumption data of similar users for all intended products. For example, the consumption amount of all intended products by similar users can be normalized, and the consumption amount of the intended products can be multiplied or added with the corresponding preference value, and the intended product matrix can be filled with the result of multiplication or addition.

[0070] It should be noted that there are usually multiple similar users to the target user. In the intended product matrix, each row represents a similar user, and each column represents the product or sum of the similar user's preference value for the intended product and the consumption amount. At the same time, in order to ensure that the number of data in each row of the intended product matrix is ​​consistent, after calculating the preference values ​​of similar users for all intended products, some intended products will be selected based on the number and preference values ​​of the intended products to construct the intended product matrix.

[0071] Optionally, the step of calculating the target user's intention level for the intended product also includes: obtaining all financial products of the financial institution, calculating the target user's preference value for all financial products based on the target user's behavioral data, and calculating the target user's active value based on all preference values; calculating the target user's intention level for similar users' intended products based on the target user's active value, the similarity value between the target user and similar users, and the intended product matrix of similar users.

[0072] After obtaining the intended product matrix, the embodiment of the present invention obtains all financial products of the financial institution, calculates the target user's preference value for all financial products based on the target user's behavior data, and then calculates the target user's active value based on all preference values. The purpose of calculating the active value is to distinguish between active users and inactive users. Active users have a relatively high desire to purchase financial products, while inactive users have a low desire to purchase financial products.

[0073] After calculating the target user's preference values ​​for all financial products, the average of the preference values ​​is calculated and used as the target user's active value. Finally, based on the target user's active value, the similarity value between the target user and similar users, and the similar user's intended product matrix, the target user's intention degree value for the similar user's intended product is calculated. The calculation formula is:

[0074]

[0075] In the above formula, r ij represents the intention degree value of target user i for similar user k’s intended product j, r kj represents the preference value of similar user k for intended product j, S ik is the similarity value between target user i and similar user k, is the average value of preference of similar user k for all intended products, is the activity value of target user i, and N(i,j) is the set of similar users of target user i.

[0076] After calculating the target user's intention level values ​​for all intended products, the target intended product for the target user is selected from all intended products according to the size of the intention level values ​​as the subsequent recommended financial product.

[0077] Optionally, the credit assessment model is constructed in the following manner: based on the historical identity data, historical asset data, historical loan data and historical credit levels of the target user and other users, a plurality of decision trees are randomly constructed using a machine learning strategy; a plurality of empirical decision trees input by an external terminal are obtained, wherein the empirical decision trees are constructed by the external terminal based on historical experience; the historical identity data, historical asset data and historical loan data of the target user and other users are respectively input into the randomly constructed decision tree and the empirical decision tree to obtain a plurality of output credit levels; based on all output credit levels and known historical credit levels, a machine learning strategy is used to determine the weights of the randomly constructed decision tree and the empirical decision tree; a credit assessment model is determined based on the randomly constructed decision tree and the corresponding weights, the empirical decision tree and the corresponding weights.

[0078] In the embodiment of the present invention, the credit assessment model is pre-built. In the process of building the credit assessment model, firstly, based on the historical identity data, historical asset data, historical loan data and historical credit level of the target user and other users, a machine learning strategy is used to randomly build multiple decision trees. When building the decision tree, Gini impurity is used as the basis for node segmentation, and Gini impurity is expressed as:

[0079]

[0080] In the above formula, C represents the classification of user D’s historical credit level, p(c|D b ) represents category c in dataset D b The probability in the dataset D b Including user D’s historical identity data, historical asset data, and historical loan data.

[0081] In addition to the randomly constructed decision trees, the embodiment of the present invention receives multiple empirical decision trees input by the external terminal. The empirical decision trees are constructed by the external terminal based on historical experience. The empirical decision trees can have higher assessment accuracy in specific credit assessment scenarios. It should be noted that the ratio of the number of random decision trees to empirical decision trees can be set by oneself. For example, random decision trees account for 80% of all decision trees, and empirical decision trees account for 20% of all decision trees.

[0082] After obtaining all decision trees, the historical identity data, historical asset data, and historical loan data of the target user and other users are respectively input into the randomly constructed decision tree and the empirical decision tree to obtain multiple output credit levels. Then, based on all output credit levels and known historical credit levels, all decision trees are iteratively trained and optimized using machine learning strategies to determine the respective weights of the randomly constructed decision tree and the empirical decision tree, and then the credit assessment model is determined.

[0083] Optionally, a pre-built credit assessment model is used to perform credit rating on a target user to obtain the target user's credit level, including: inputting the target user's identity data, asset data, and loan data into multiple decision trees in the credit assessment model respectively to obtain multiple output credit levels; and determining the target user's credit level based on the weights and output credit levels of all decision trees.

[0084] After the credit assessment model is determined, it is used in the actual scenario to rate the credit of the target user. When using the credit assessment model to rate the credit of the target user, the identity data, asset data, and loan data of the target user are respectively input into multiple decision trees in the credit assessment model to obtain multiple output credit levels. The credit level of the target user is determined based on the weights and output credit levels of all decision trees. The calculation formula for determining the credit level of the target user is:

[0085]

[0086] In the above formula, J is the final credit level of the target user, b represents the sequence number of the decision tree, B represents the number of decision trees, x represents the input data of the decision tree, and T b (x) represents the output credit level of the bth decision tree, c represents the credit level, T b (x) = c represents the output credit level of the bth decision tree is credit level c, F is the indicator function, φ is the decision tree weight, argmax c Indicates the choice The largest category is c.

[0087] This calculation formula represents the weighted integration of the output results of all decision trees, and determines a final credit level based on the output credit level of each decision tree and the weight of each decision tree. This process ensures that the final credit level not only takes into account the diversity of credit assessment models, but also reflects the contribution of each decision tree to the final credit level, making the final credit level highly accurate.

[0088] Optionally, in addition to sending recommendation information to the target user, it also includes: sending account change information, subscription service notifications, event promotion information and market information to the target user based on the target user's behavior data, consumption data and target intended products.

[0089] In the embodiment of the present invention, in addition to sending recommendation information to the target user, the target user will also be sent account change information, subscription service notifications, activity promotion information, and market information based on the target user's behavior data, consumption data, and target intended products. For example, when the target user purchases a financial product, the embodiment of the present invention sends the purchase information and financial account change information to the target user to notify the target user of the successful purchase.

[0090] Figure 3 is a simplified flow chart of an optional artificial intelligence-based financial product recommendation method according to an embodiment of the present invention, such as Figure 3 As shown, after completing the identity authentication of the target user, the multi-source data of the target user and other users is obtained and preprocessed. Based on the preprocessed multi-source data, a machine learning strategy is used to generate financial products that the target user is interested in, that is, target intended products. The target user is then credit evaluated through a credit assessment model to generate the target user's credit level. Finally, based on the target user's target intended products and credit level, personalized recommendation information is sent to the target user.

[0091] The embodiment of the present invention obtains multi-source data of the target user and other users and pre-processes the multi-source data, thereby effectively integrating data from different sources and providing comprehensive data support for the subsequent generation of target intended products and user credit evaluation.

[0092] By calculating the similarity value between the target user and other users based on multi-source data, and then calculating the target user's intention level for financial products based on the similarity value and the behavioral data and consumption data of similar users on the intended products, the target user's target intended products are determined, which better matches the target user's asset status and preference characteristics, and provides users with more accurate and personalized financial product recommendations.

[0093] At the same time, by using a pre-built credit assessment model to rate the target users, we obtained a precise target user credit level, and combined the user's identity data, asset data and loan data to accurately assess the user's credit level. By combining intelligently generated financial products with accurate credit levels to recommend financial products to users, we better matched the user's preference characteristics and credit characteristics, effectively improving the accuracy and personalization of financial product recommendations, and thus solving the technical problems in related technologies such as low matching of recommendation algorithms, low accuracy of user credit assessment results, and low personalization of recommended financial products.

[0094] The following is a detailed description in conjunction with another embodiment.

[0095] Embodiment 2

[0096] An artificial intelligence-based financial product recommendation system provided in this embodiment includes multiple implementation units, each of which corresponds to each implementation step in the above-mentioned embodiment 1.

[0097] Figure 4 is a schematic diagram of an optional artificial intelligence-based financial product recommendation system according to an embodiment of the present invention, such as Figure 4 As shown, the artificial intelligence-based financial product recommendation system may include: a user authentication unit 41, a data management unit 42, an asset management unit 43, a user credit evaluation unit 44, and a message push unit 45.

[0098] The user authentication unit 41 is used to perform real-name authentication on the target user and other users, and to establish connections with the associated accounts of the target user and other users.

[0099] The data management unit 42 is used to manage the multi-source data of the target user and other users, and to pre-process the multi-source data.

[0100] The asset management unit 43 is used to calculate the similarity value between the target user and other users, determine the similar users of the target user, and calculate the target user's intention level value for the intended product based on the similarity value and the behavioral data and consumption data of the similar users for the intended product, and determine the target intended product of the target user from the intended products according to the intention level value.

[0101] The user credit evaluation unit 44 is used to perform credit rating on the target user according to the identity data, asset data and loan data of the target user using a pre-built credit evaluation model to obtain the credit level of the target user.

[0102] The message push unit 45 is used to generate product recommendation information according to the target intended product and credit level of the target user, and send the product recommendation information to the client held by the target user.

[0103] The above-mentioned artificial intelligence-based financial product recommendation system obtains multi-source data of the target user and other users through the data management unit 42, pre-processes the multi-source data, and then calculates the similarity value between the target user and other users based on the multi-source data through the asset management unit 43, determines similar users based on the similarity value, and defines the financial products browsed or purchased by similar users as intended products, and then calculates the target user's intention level value for the intended product based on the similarity value and the behavioral data and consumption data of similar users on the intended product, selects the target intended product for the target user from all intended products according to the intention level value, and then uses the user credit evaluation unit 44 to credit rate the target user based on the target user's identity data, asset data and loan data using a pre-built credit evaluation model to obtain the target user's credit level, and finally generates product recommendation information based on the target user's target intended product and credit level through the message push unit 45, and sends the product recommendation information to the client held by the target user.

[0104] The embodiments of the present invention acquire multi-source data and pre-process the data, thereby effectively integrating data from different sources, providing comprehensive data support for the recommendation of financial products and the credit evaluation of users, and generating recommended financial products by calculating the similarity values ​​between users and based on the user's behavior data and consumption data. At the same time, by using a pre-built credit evaluation model and combining the user's identity data, asset data and loan data, an accurate credit level evaluation of the user is performed, and financial products are recommended to the user by combining intelligently generated financial products with accurate credit levels, thereby better matching the user's preference characteristics and credit characteristics, and effectively improving the accuracy and personalization of financial product recommendations, thereby solving the technical problems in the related art of low matching of recommendation algorithms, low accuracy of user credit evaluation results, and low personalization of recommended financial products.

[0105] Optionally, the user authentication unit 41 includes: a real-name authentication module, used to perform real-name authentication on the target user and other users, the real-name authentication operations include: identity information verification and biometric verification, wherein identity information verification includes: name, identification document verification, biometric verification includes: fingerprint, face verification; an account connection module, used to establish a connection with the target user and other users' associated accounts, wherein the associated accounts include: financial institution accounts, e-commerce platform accounts and social platform accounts.

[0106] Optionally, the data management unit 42 includes: a management module for managing multi-source data of the target user and other users; and a pre-processing module for pre-processing the multi-source data.

[0107] Optionally, the asset management unit 43 includes: a first calculation module, which calculates the similarity value between the target user and other users based on multi-source data, and determines similar users based on the similarity value; a definition module, which is used to define financial products browsed or purchased by similar users as intended products; a scoring module, which is used to obtain the behavioral data of similar users for all intended products, and score the behavioral data based on preset rules; a second calculation module, which is used to obtain the time point when the similar user generates the behavioral data, and calculate the time difference between the occurrence time point of the behavioral data and the current time point; a third calculation module, for each intended product, based on the score and time difference of the corresponding behavioral data, calculates the preference value of the similar user for the intended product; a first construction module, which constructs an intended product matrix based on the preference values ​​of similar users for all intended products and the consumption data of the intended products.

[0108] Optionally, the asset management unit 43 also includes: a fourth calculation module, which is used to obtain all financial products of the financial institution, calculate the target user's preference value for all financial products based on the target user's behavioral data, and calculate the target user's active value based on all preference values; a fifth calculation module, which calculates the target user's intention level value for similar users' intended products based on the target user's active value, the similarity value between the target user and similar users, and the similar users' intended product matrix, and determines the target user's target intended product from the intended products according to the intention level value.

[0109] Optionally, the user credit assessment unit 44 includes: a second construction module, which randomly constructs multiple decision trees based on historical identity data, historical asset data, historical loan data and historical credit levels of the target user and other users using a machine learning strategy; an acquisition module, which is used to obtain multiple experience decision trees input by an external terminal, wherein the experience decision trees are constructed by the external terminal based on historical experience; a first input module, which is used to input the historical identity data, historical asset data and historical loan data of the target user and other users into the randomly constructed decision trees and the experience decision trees respectively, to obtain multiple output credit levels; a first determination module, which determines the weights of the randomly constructed decision trees and the experience decision trees based on all output credit levels and known historical credit levels using a machine learning strategy; a second determination module, which determines the credit assessment model based on the randomly constructed decision trees and corresponding weights, the experience decision trees and corresponding weights.

[0110] Optionally, the user credit assessment unit 44 also includes: a second input module, which is used to input the identity data, asset data and loan data of the target user into multiple decision trees in the credit assessment model respectively to obtain multiple output credit levels; and a third determination module, which determines the credit level of the target user based on the weights and output credit levels of all decision trees.

[0111] Optionally, the message push unit 45 includes: a first push module, which is used to generate product recommendation information based on the target user's target intended products and credit level, and send the product recommendation information to the client held by the target user; a second push module, which sends account change information, subscription service notifications, event promotion information and market information information to the target user based on the target user's behavior data, consumption data and target intended products.

[0112] The above-mentioned artificial intelligence-based financial product recommendation system may also include a processor and a memory. The above-mentioned user authentication unit 41, data management unit 42, asset management unit 43, user credit assessment unit 44, message push unit 45, etc. are all stored in the memory as program units, and the processor executes the above-mentioned program units stored in the memory to realize corresponding functions.

[0113] The processor includes a kernel, which retrieves the corresponding program unit from the memory. One or more kernels can be set, and financial products can be recommended to target users based on artificial intelligence by adjusting kernel parameters.

[0114] The above-mentioned memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one storage chip.

[0115] Embodiment 3

[0116] An embodiment of the present application may provide an electronic device, Figure 5 is a structural block diagram of an electronic device according to an embodiment of the present application. Figure 5 As shown, the electronic device may include: one or more ( Figure 5 (only one is shown) processor 502, memory 504, storage controller, and peripheral interface, wherein the peripheral interface is connected to the radio frequency module, audio module and display.

[0117] Among them, the memory can be used to store software programs and modules, such as program instructions / modules corresponding to the methods and devices in the embodiments of the present application, and the processor executes various functional applications and data processing by running the software programs and modules stored in the memory, that is, realizing the above-mentioned method. The memory may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory may further include a memory remotely arranged relative to the processor, and these remote memories may be connected to the terminal via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0118] The processor can call the information and application stored in the memory through the transmission device to perform the following steps: obtain multi-source data of the target user and other users, and pre-process the multi-source data, wherein the other users are customers of financial institutions other than the target user, and the multi-source data include: identity data, behavior data, consumption data, asset data and loan data; based on the multi-source data, calculate the similarity value between the target user and other users, determine similar users based on the similarity value, and define the financial products browsed or purchased by similar users as intended products; based on the similarity value and the behavior data and consumption data of similar users on the intended products, calculate the target user's intention level value for the intended products, and select the target intended product for the target user from all intended products according to the intention level value; based on the target user's identity data, asset data and loan data, use a pre-built credit assessment model to credit rate the target user to obtain the target user's credit level; based on the target user's target intended product and credit level, generate product recommendation information, and send the product recommendation information to the client held by the target user.

[0119] The processor can also call the information and applications stored in the memory through the transmission device to perform the following steps: perform real-name authentication on the target user and other users, the real-name authentication operations include: identity information verification and biometric verification, wherein identity information verification includes: name, identification document verification, biometric verification includes: fingerprint, face verification; establish connection with the associated accounts of the target user and other users, wherein the associated accounts include: financial institution accounts, e-commerce platform accounts and social platform accounts.

[0120] The processor can also call the information and application programs stored in the memory through the transmission device to perform the following steps: obtain the behavioral data of similar users for all intended products, and score the behavioral data based on preset rules; obtain the time points when similar users generate behavioral data, and calculate the time difference between the time point when the behavioral data occurs and the current time point; for each intended product, calculate the preference value of similar users for the intended product based on the score and time difference of the corresponding behavioral data; construct an intended product matrix based on the preference values ​​of similar users for all intended products and the consumption data of the intended products.

[0121] The processor can also call the information and application programs stored in the memory through the transmission device to perform the following steps: obtain all financial products of the financial institution, calculate the target user's preference value for all financial products based on the target user's behavioral data, and calculate the target user's active value based on all preference values; calculate the target user's intention level value for similar users' intended products based on the target user's active value, the similarity value between the target user and similar users, and the intended product matrix of similar users.

[0122] The processor can also call the information and application programs stored in the memory through the transmission device to perform the following steps: based on the historical identity data, historical asset data, historical loan data and historical credit levels of the target user and other users, randomly construct multiple decision trees using a machine learning strategy; obtain multiple experience decision trees input by an external terminal, wherein the experience decision tree is constructed by the external terminal based on historical experience; input the historical identity data, historical asset data and historical loan data of the target user and other users into the randomly constructed decision tree and the experience decision tree respectively to obtain multiple output credit levels; based on all output credit levels and known historical credit levels, use a machine learning strategy to determine the weights of the randomly constructed decision tree and the experience decision tree; determine the credit assessment model based on the randomly constructed decision tree and the corresponding weights, the experience decision tree and the corresponding weights.

[0123] The processor can also call the information and application programs stored in the memory through the transmission device to perform the following steps: input the identity data, asset data and loan data of the target user into multiple decision trees in the credit assessment model respectively to obtain multiple output credit levels; determine the credit level of the target user based on the weights and output credit levels of all decision trees.

[0124] The processor can also call the information and applications stored in the memory through the transmission device to perform the following steps: based on the target user's behavior data, consumption data and target intended products, send account change information, subscription service notifications, event promotion information and market information information to the target user.

[0125] By using the embodiment of the present application, a solution for a financial product recommendation method based on artificial intelligence is provided. By obtaining multi-source data of the target user and other users, pre-processing the multi-source data, and then calculating the similarity value between the target user and other users based on the multi-source data, similar users are determined based on the similarity value, and financial products browsed or purchased by similar users are defined as intended products. Then, based on the similarity value and the behavioral data and consumption data of similar users on the intended products, the target user's intention degree value for the intended product is calculated, and the target intended product for the target user is selected from all intended products according to the intention degree value. Then, based on the target user's identity data, asset data and loan data, a pre-built credit assessment model is used to perform credit rating on the target user to obtain the target user's credit level. Finally, based on the target user's target intended product and credit level, product recommendation information is generated, and the product recommendation information is sent to the client held by the target user, thereby achieving the purpose of effectively integrating data from different sources, accurately assessing the user's credit level, and better matching the user's preference characteristics and credit characteristics, and effectively improving the accuracy and personalization of financial product recommendations, thereby solving the technical problems in related technologies such as low matching degree of recommendation algorithms and low accuracy of user credit assessment results, resulting in low personalization of recommended financial products.

[0126] It can be understood by those skilled in the art that Figure 5 The structure shown is for illustration only, and the electronic device may also be a smart phone, a tablet computer, a PDA, a mobile Internet device (MID), a PAD, or other terminal devices. Figure 5 The structure of the electronic device is not limited. Figure 5 More or fewer components (such as network interfaces, display devices, etc.) shown in, or having Figure 5 Different configurations are shown.

[0127] A person of ordinary skill in the art can understand that all or part of the steps in the various artificial intelligence-based financial product recommendation methods in the above-mentioned embodiments can be completed by instructing the hardware related to the terminal device through a program, and the program can be stored in a computer-readable storage medium, and the storage medium may include: a flash drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, etc.

[0128] Embodiment 4

[0129] The embodiment of the present application further provides a storage medium. Optionally, in this embodiment, the storage medium can be used to store the program code executed by the artificial intelligence-based financial product recommendation method provided in the first embodiment.

[0130] According to another aspect of an embodiment of the present invention, a computer-readable storage medium is also provided, the computer-readable storage medium including a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute any one of the artificial intelligence-based financial product recommendation methods in the above-mentioned embodiment 1.

[0131] Optionally, in this embodiment, the above storage medium may be located in any computer terminal in a computer terminal group in a computer network, or in any mobile terminal in a mobile terminal group.

[0132] The present application also provides a computer program product, including a computer program, which, when executed by a processor, implements the steps of the artificial intelligence-based financial product recommendation method described in each embodiment of the present application.

[0133] The present application also provides a computer program product, including a non-volatile computer-readable storage medium, wherein the non-volatile computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the artificial intelligence-based financial product recommendation method described in each embodiment of the present application are implemented.

[0134] The serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.

[0135] In the above embodiments of the present invention, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0136] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only schematic. For example, the division of the units can be a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0137] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0138] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0139] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk and other media that can store program codes.

[0140] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A financial product recommendation method based on artificial intelligence, characterized in that: include: Acquire multi-source data of the target user and other users, and pre-process the multi-source data, wherein the other users are customers of financial institutions other than the target user, and the multi-source data includes: identity data, behavior data, consumption data, asset data, and loan data; Based on the multi-source data, a similarity value between the target user and the other users is calculated, similar users are determined based on the similarity value, and financial products browsed or purchased by the similar users are defined as intended products; Based on the similarity value and the behavior data and consumption data of the similar users on the intended product, calculating the intention degree value of the target user for the intended product, and selecting a target intended product for the target user from all the intended products according to the intention degree value; Based on the identity data, asset data and loan data of the target user, a pre-built credit assessment model is used to perform a credit rating on the target user to obtain the credit level of the target user; Based on the target intended product and the credit level of the target user, product recommendation information is generated, and the product recommendation information is sent to a client held by the target user.

2. The financial product recommendation method according to claim 1, characterized in that: Before acquiring the multi-source data of the target user and the other users, the method includes: Performing real-name authentication on the target user and the other users, wherein the real-name authentication operation includes: identity information verification and biometric verification, wherein the identity information verification includes: name and identity document verification, and the biometric verification includes: fingerprint and face verification; Establish connections with the target user and associated accounts of other users, wherein the associated accounts include: financial institution accounts, e-commerce platform accounts, and social platform accounts.

3. The financial product recommendation method according to claim 1, characterized in that: The step of calculating the target user's intention level value for the intended product includes: Obtaining the behavior data of the similar users on all the intended products, and scoring the behavior data based on preset rules; Obtain the time point at which the similar user generates the behavior data, and calculate the time difference between the time point at which the behavior data occurs and the current time point; For each of the intended products, based on the score of the corresponding behavior data and the time difference, calculate the preference value of the similar user for the intended product; Based on the preference values ​​of the similar users for all the intended products and the consumption data of the intended products, an intended product matrix is ​​constructed.

4. The financial product recommendation method according to claim 3, characterized in that: The step of calculating the target user's intention level value for the intended product also includes: Obtain all financial products of the financial institution, calculate the target user's preference values ​​for all financial products based on the target user's behavior data, and calculate the target user's activity value based on all preference values; Based on the activity value of the target user, the similarity value between the target user and the similar users, and the intended product matrix of the similar users, the intention degree value of the target user for the intended products of the similar users is calculated.

5. The financial product recommendation method according to claim 1, characterized in that: The credit assessment model is constructed in the following way: Based on the historical identity data, historical asset data, historical loan data, and historical credit rating of the target user and the other users, a machine learning strategy is used to randomly construct multiple decision trees; Acquire multiple experience decision trees input by an external terminal, wherein the experience decision trees are constructed by the external terminal based on historical experience; Inputting the historical identity data, historical asset data and historical loan data of the target user and the other users into the randomly constructed decision tree and the experience decision tree respectively to obtain multiple output credit levels; Based on all the output credit levels and the known historical credit levels, a machine learning strategy is used to determine the weights of the randomly constructed decision tree and the empirical decision tree; The credit assessment model is determined based on the randomly constructed decision tree and the corresponding weights, the empirical decision tree and the corresponding weights.

6. The financial product recommendation method according to claim 1, characterized in that: The step of using the pre-built credit assessment model to perform credit rating on the target user to obtain the credit level of the target user includes: Inputting the identity data, asset data and loan data of the target user into multiple decision trees in the credit assessment model respectively to obtain multiple output credit levels; Based on the weights and output credit levels of all decision trees, the credit level of the target user is determined.

7. The financial product recommendation method according to claim 1, characterized in that: In addition to sending the recommendation information to the target user, the method further includes: Based on the target user's behavior data, consumption data and target intended products, account change information, subscription service notifications, event promotion information and market information are sent to the target user.

8. A financial product recommendation system based on artificial intelligence, characterized in that: include: A user authentication unit, used to perform real-name authentication on a target user and other users, and to establish a connection with the associated accounts of the target user and other users; A data management unit, used for managing the multi-source data of the target user and the other users, and pre-processing the multi-source data; an asset management unit, configured to calculate a similarity value between the target user and the other users, determine similar users of the target user, and calculate a value of the target user's intention level for the intended product based on the similarity value and the behavior data and consumption data of the similar users for the intended product, and determine a target intended product of the target user from the intended products according to the intention level value; A user credit evaluation unit, configured to perform a credit rating on the target user using a pre-built credit evaluation model according to the identity data, asset data and loan data of the target user, so as to obtain the credit level of the target user; A message push unit is used to generate product recommendation information according to the target intended product and the credit level of the target user, and send the product recommendation information to the client held by the target user.

9. An electronic device, characterized in that: It includes one or more processors and a memory, wherein the memory is used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the artificial intelligence-based financial product recommendation method described in any one of claims 1 to 7.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the artificial intelligence-based financial product recommendation method described in any one of claims 1 to 7 are implemented.

Images