Financial product recommendation method and device based on artificial intelligence, equipment and medium

Abstract

The invention provides a financial product recommendation method and device based on artificial intelligence, equipment and a medium, and belongs to the field of artificial intelligence, and the method comprises the steps: obtaining a user portrait of a customer who selects and purchases a financial product and real-time market data of the financial product; performing product combination on a plurality of financial products based on the user portrait and the real-time market data through a preset product combination model to generate a target financial product library; and through a preset product recommendation model, based on the user portrait and the target financial product library, carrying out financial product recommendation to obtain a financial product recommendation list of the customer who selects and purchases the financial product. According to the method, the financial product recommendation list matched with the customer purchasing the financial product can be accurately obtained, the efficiency and accuracy of financial product recommendation are greatly improved, and then the recognition degree and satisfaction degree of the customer to the financial product are greatly improved. The method can be applied to the technical field of finance, and the accuracy of financial product recommendation is improved.

Description

Technical Field

[0001] This application relates to the field of artificial intelligence technology, and in particular to a method, apparatus, device, and medium for recommending financial products based on artificial intelligence. Background Technology

[0002] Financial products refer to various vehicles used in the financial process, including currencies, gold, foreign exchange, securities, and insurance. Financial products are the objects of buying and selling in the financial market. Supply and demand, through market competition, determine the price of financial products, such as interest rates or yields, ultimately completing the transaction and achieving the goal of financing. Currently, financial product recommendation strategies mainly rely on historical data, fixed rules, and manual analysis. However, current methods suffer from low matching rates and poor customer satisfaction, leading to high customer churn rates.

[0003] Therefore, how to accurately recommend financial products to customers in order to improve customer satisfaction is an urgent problem to be solved. Summary of the Invention

[0004] The main purpose of this application is to provide a method, apparatus, device, and medium for recommending financial products based on artificial intelligence, aiming to improve the efficiency and accuracy of financial product recommendations.

[0005] Firstly, this application provides a method for recommending financial products based on artificial intelligence, which includes the following steps:

[0006] Acquire user profiles of customers who purchase financial products and real-time market data of financial products, including real-time sales data, yields and market risk values ​​of various types of financial products.

[0007] Based on the user profile and the real-time market data, multiple financial products are combined using a preset product combination model to generate a target financial product library. The target financial product library includes single financial products and financial product groups of multiple financial products. The preset product combination model is obtained by training a preset neural network model based on multiple sample data.

[0008] Based on the user profile and the target financial product library, a preset product recommendation model is used to recommend financial products, thereby obtaining a list of recommended financial products for the customer who has selected financial products. The preset product recommendation model is obtained by training a preset neural network model based on multiple sample data.

[0009] Secondly, this application also provides a financial product recommendation device, which includes an acquisition module and a generation module, wherein:

[0010] The acquisition module is used to acquire user profiles of customers who purchase financial products and real-time market data of financial products. The real-time market data includes real-time sales data, yields, and market risk values ​​of various types of financial products.

[0011] The generation module is used to combine multiple financial products based on the user profile and the real-time market data using a preset product combination model to generate a target financial product library. The target financial product library includes single financial products and financial product groups of multiple financial products. The preset product combination model is obtained by training a preset neural network model based on multiple sample data in advance.

[0012] The generation module is further configured to recommend financial products from the target financial product library based on the user profile using a preset product recommendation model, thereby obtaining a list of financial product recommendations for the customer who selects financial products. The preset product recommendation model is obtained by training a preset neural network model based on multiple sample data.

[0013] Thirdly, this application also provides a computer device, the computer device including a processor, a memory, and a computer program stored in the memory and executable by the processor, wherein when the computer program is executed by the processor, it implements the steps of the artificial intelligence-based financial product recommendation method described above.

[0014] Fourthly, this application also provides a computer-readable storage medium storing a computer program, wherein when the computer program is executed by a processor, it implements the steps of the artificial intelligence-based financial product recommendation method described above.

[0015] This application provides a method, apparatus, device, and medium for recommending financial products based on artificial intelligence. The application obtains user profiles of customers selecting financial products and real-time market data of the financial products. This real-time market data includes real-time sales data, yield rates, and market risk values ​​for various types of financial products. Based on the user profiles and the real-time market data, a preset product combination model is used to combine multiple financial products to generate a target financial product library. This target financial product library includes single financial products and groups of multiple financial products. The preset product combination model is obtained by pre-training a preset neural network model based on multiple sample data. Based on the user profiles and the target financial product library, a preset product recommendation model is used to recommend financial products to customers selecting financial products, resulting in a recommended list of financial products for them. This preset product recommendation model is also obtained by pre-training a preset neural network model based on multiple sample data. The product portfolio model in this application combines multiple financial products based on user profiles and real-time market data, which can accurately obtain a target financial product library that matches the market in real time. The product recommendation model recommends financial products from the target financial product library based on user profiles, which can accurately obtain a list of financial product recommendations that match the customers who purchase financial products, greatly improving the efficiency and accuracy of financial product recommendations, and thus greatly improving customers' recognition and satisfaction with financial products. Attached Figure Description

[0016] To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 A flowchart illustrating an artificial intelligence-based financial product recommendation method provided in this application embodiment;

[0018] Figure 2 A flowchart illustrating another AI-based financial product recommendation method provided in this application embodiment;

[0019] Figure 3 This is a schematic diagram of the structure of a product portfolio model provided in an embodiment of this application;

[0020] Figure 4 for Figure 1 A flowchart illustrating the sub-steps of an AI-based financial product recommendation method.

[0021] Figure 5A schematic block diagram of a financial product recommendation device provided in this application embodiment;

[0022] Figure 6 This is a schematic block diagram of the structure of a computer device provided in an embodiment of this application.

[0023] The realization of the purpose, functional features and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0024] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0025] The flowchart shown in the attached diagram is for illustrative purposes only and does not necessarily include all content and operations / steps, nor does it necessarily have to be performed in the order described. For example, some operations / steps can be broken down, combined, or partially merged, so the actual execution order may change depending on the actual situation.

[0026] The embodiments of this application can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence (AI) refers to the theories, methods, technologies, and application systems that use digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0027] Foundational technologies for artificial intelligence generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing, operating / interactive systems, and mechatronics. AI software technologies mainly encompass computer vision, robotics, biometrics, speech processing, natural language processing, and machine learning / deep learning.

[0028] Financial products refer to various vehicles used in the financial process, including currencies, gold, foreign exchange, securities, and insurance. Financial products are the objects of buying and selling in the financial market. Supply and demand, through market competition, determine the price of financial products, such as interest rates or yields, ultimately completing the transaction and achieving the goal of financing. Currently, financial product recommendation strategies mainly rely on historical data, fixed rules, and manual analysis. However, current methods suffer from low matching rates and poor customer satisfaction, leading to high customer churn rates.

[0029] To address the aforementioned issues, this application provides an AI-based financial product recommendation method, apparatus, device, and medium. The AI-based financial product recommendation method includes: acquiring user profiles of customers selecting financial products and real-time market data for the financial products, wherein the real-time market data includes real-time sales data, yield rates, and market risk values ​​for various types of financial products; combining multiple financial products based on the user profiles and the real-time market data using a preset product combination model to generate a target financial product library, wherein the target financial product library includes single financial products and groups of multiple financial products, and the preset product combination model is obtained by pre-training a preset neural network model based on multiple sample data; and recommending financial products based on the user profiles and the target financial product library using a preset product recommendation model to obtain a financial product recommendation list for the customers selecting financial products, wherein the preset product recommendation model is obtained by pre-training a preset neural network model based on multiple sample data.

[0030] The AI-based financial product recommendation method can be applied to computer devices, such as mobile phones, tablets, laptops, desktop computers, personal digital assistants, and wearable devices.

[0031] The following detailed description of some embodiments of this application is provided in conjunction with the accompanying drawings. Unless otherwise specified, the following embodiments and features can be combined with each other.

[0032] Please refer to Figure 1 , Figure 1 This is a flowchart illustrating an artificial intelligence-based financial product recommendation method provided for an embodiment of this application.

[0033] like Figure 1 As shown, the AI-based financial product recommendation method includes steps S101 to S103.

[0034] Step S101: Obtain user profiles of customers who purchase financial products and real-time market data of financial products.

[0035] The user profile is data linked together from user attributes and behavioral data, while the real-time market data for the financial product includes the product's coverage, profit margin, and market risk, etc.

[0036] In some embodiments, asset data, consumption behavior data, and health data of customers who purchase financial products are acquired; user profiles of these customers are constructed based on the asset data, consumption behavior data, and health data. Accurate user profiles of customers who purchase financial products can be constructed based on the asset data, consumption behavior data, and health data.

[0037] In some embodiments, the asset data of customers who purchase financial products can be obtained by: obtaining the account balances and investment data of customers who purchase financial products in various financial institutions, thereby obtaining the asset data of customers who purchase financial products.

[0038] In some embodiments, the consumption behavior data of customers who purchase financial products can be obtained by: obtaining the transaction flow information of customers who purchase financial products to obtain the consumption behavior data of customers who purchase financial products.

[0039] In some embodiments, the health data of customers who purchase financial products can be obtained by: obtaining medical consultation data of customers who purchase financial products, and / or monitoring data collected by wearable devices worn by customers who purchase financial products, to obtain the health data of customers who purchase financial products.

[0040] It should be noted that the asset data, consumption behavior data, and health data of customers who purchase financial products are all encrypted using a preset encryption algorithm, which effectively ensures user data security. The preset encryption algorithm can be selected according to actual needs, and this application embodiment does not impose a specific limitation on it. For example, the preset encryption algorithm can be the homomorphic encryption algorithm Paillier. Building user profiles is a conventional technical means, and this application will not elaborate on it further.

[0041] In some embodiments, real-time sales data, yield, and market risk value of each type of financial product are acquired to obtain real-time market data for each type of financial product. This real-time market data for each type of financial product is then used as the real-time market data for each financial product within that type. By statistically analyzing the real-time market data for each type and using the real-time market data for each type of financial product as the real-time market data for each financial product within that type, the efficiency of acquiring real-time market data for financial products can be effectively improved.

[0042] The real-time sales data refers to the sales volume of financial products within a preset period, the rate of return refers to the percentage of the net profit of the financial products within the preset period to the average principal used, and the market risk value refers to the risk fluctuation value of the financial products within the preset period. The preset period can be set according to the actual situation, and this application embodiment does not make specific limitations on it. For example, the preset period can be set to 1 week.

[0043] Step S102: Based on the user profile and the real-time market data, multiple financial products are combined using a preset product combination model to generate a target financial product library.

[0044] The target financial product library includes single financial products and groups of financial products. For example, a group of financial products may include financial product 1, financial product 2, and financial product 3. This pre-defined product portfolio model is obtained by training a pre-defined neural network model based on multiple sample data.

[0045] In some embodiments, such as Figure 2 As shown, the AI-based financial product recommendation method also includes steps S201 to S205.

[0046] Step S201: Obtain the first sample dataset.

[0047] The first sample dataset includes multiple first sample data sets, which include sample user profiles, sample real-time market data, and labeled target financial product databases.

[0048] In some embodiments, a user profile of a customer selecting financial products is obtained and labeled as a sample user profile. Real-time market data for multiple financial products are obtained and labeled as sample real-time market data for each financial product. Based on the user profile and real-time market data, products are combined to obtain multiple financial products and / or multiple financial product groups. These multiple financial products and / or multiple financial product groups are labeled to obtain a labeled target financial product library. The sample user profile, sample real-time market data, and labeled target financial product library are used as a single sample dataset. The aforementioned steps are repeated to obtain a sample dataset.

[0049] Step S202: Obtain a preset neural network model and select a sample data from the first sample dataset as the target sample data.

[0050] This neural network model includes, but is not limited to, convolutional neural network models, recurrent neural network models, and recurrent convolutional neural network models. For example... Figure 3 As shown, the preset neural network model includes an encoder layer, a fully connected layer, and a decoder layer. The encoder layer is used to encode features of user profiles and real-time market data. The fully connected layer is used to predict product portfolios of multiple financial products. The decoder layer is used to decode the feature vectors of financial product portfolios.

[0051] In some embodiments, a sample data point is selected from a first sample dataset as the target sample data to accurately obtain the target sample data. This first target sample data includes a sample user profile, sample real-time market data, and a labeled target financial product database.

[0052] Step S203: Based on the sample user profiles and sample real-time market data in the target sample data, a preset neural network model is used to combine multiple financial products to generate a predicted target financial product library.

[0053] The sample user profile and real-time market data are input into a pre-defined neural network model to combine multiple financial products, generating a predicted target financial product library. By combining sample user profiles and real-time market data using the pre-defined neural network model, the predicted target financial product library can be accurately obtained.

[0054] In some embodiments, the encoder layer encodes the sample user profile and the sample real-time market data to obtain the predicted user feature vector and the predicted real-time market data feature vector, respectively; the fully connected layer performs product portfolio prediction on multiple financial products based on the predicted user feature vector and the predicted real-time market data feature vector to obtain the predicted financial product portfolio feature vector; and the decoder layer decodes the predicted financial product portfolio feature vector to obtain the predicted target financial product library.

[0055] In some embodiments, the method for obtaining the predicted financial product portfolio feature vector by performing product portfolio prediction on multiple financial products through a fully connected layer based on the predicted user feature vector and the predicted real-time market data feature vector can be as follows: Based on the predicted real-time market data feature vector of each financial product, determine the market value feature vector of each predicted financial product portfolio, which is determined based on the predicted return and market risk value of the financial product portfolio; determine the matching degree between the customer purchasing the financial product and each predicted financial product portfolio based on the market value feature vector of each predicted financial product portfolio and the predicted user feature vector; perform vector fusion on the feature vectors of the predicted financial product portfolios with matching degrees greater than or equal to a preset matching value to obtain the predicted financial product portfolio feature vector. The preset matching value can be set according to actual conditions, and this embodiment does not specifically limit it; for example, the preset matching value can be set to 0.9.

[0056] Step S204: Based on the predicted target financial product library and the labeled target financial product library, determine whether the preset neural network model has converged.

[0057] Based on the predicted and labeled target financial product databases, a loss value for a preset neural network model is determined. If the loss value of the preset neural network model is less than or equal to the preset loss value, the preset neural network model is determined to have converged; if the loss value of the preset neural network model is greater than the preset loss value, the preset neural network model is determined to have not converged. The preset loss value can be set according to actual conditions, and this embodiment does not impose specific limitations on it. For example, the preset loss value can be set to 0.02. Based on the predicted and labeled target financial product databases, the loss value of the preset neural network model can be accurately determined, and based on this loss value, it is possible to accurately determine whether the preset neural network model has converged.

[0058] In some embodiments, the method for determining the loss value of a preset neural network model based on the predicted target financial product library and the labeled target financial product library can be as follows: calculate the similarity between the predicted target financial product library and the labeled target financial product library to obtain the current similarity; obtain the historical similarity, which is the average of the current similarities of each sample data that has been trained; calculate the average of the current similarity and the historical similarity to obtain the target similarity; subtract the target similarity from the unit value to determine the loss value. The method for calculating the similarity can be selected according to the actual situation, and this application embodiment does not specifically limit it. For example, the method for calculating the similarity can be to calculate the cosine similarity between the predicted target financial product library and the labeled target financial product library.

[0059] Step S205: If the preset neural network model does not converge, the model parameters of the preset neural network model are adjusted, and the process of selecting a sample data from the first sample dataset as the target sample data continues until a converged product portfolio model is obtained.

[0060] If the loss value of the preset neural network model is greater than the preset loss value, it is determined that the preset neural network model has not converged; the model parameters of the preset neural network model are adjusted, and the process of selecting a sample data from the first sample dataset as the target sample data continues; the preset neural network model combines multiple financial products based on the sample user profile and sample real-time market data in the target sample data to generate a predicted target financial product library; the process of determining whether the preset neural network model has converged based on the predicted target financial product library and the labeled target financial product library continues until a converged product combination model is obtained.

[0061] In some embodiments, such as Figure 4 As shown, step S102 includes sub-steps S1021 to S1023.

[0062] Sub-step S1021: The encoder layer performs feature encoding on the user profile and the real-time market data respectively to obtain the user feature vector and the real-time market data feature vector.

[0063] In some embodiments, the encoder layer encodes the user profile to obtain a user feature vector; and the encoder layer also encodes the real-time market data to obtain a real-time market data feature vector. By encoding the user profile and the real-time market data separately through the encoder layer, the user feature vector and the real-time market data feature vector can be accurately obtained, greatly improving the efficiency and accuracy of financial product recommendations.

[0064] Sub-step S1022: The fully connected layer performs product portfolio prediction on multiple financial products based on the user feature vector and the real-time market data feature vector to obtain the financial product portfolio feature vector.

[0065] Based on the real-time market data feature vector of each financial product, a market value feature vector for each financial product portfolio is determined. This market value feature vector is determined based on the return rate and market risk value of the financial product portfolio. Based on the market value feature vector and user feature vector of each financial product portfolio, the matching degree between the customer purchasing the financial products and each financial product portfolio is determined. Feature vectors of financial product portfolios with matching degrees greater than or equal to a preset matching value are fused to obtain the financial product portfolio feature vector. The preset matching value can be set according to actual conditions, and this embodiment does not impose specific limitations on it. For example, the preset matching value can be set to 0.9.

[0066] In some embodiments, the market value feature vector of each financial product portfolio can be determined based on the real-time market data feature vector of each financial product by fusing the yield feature vector and the market risk value feature vector of each financial product in the financial product portfolio to obtain the market value feature vector of the financial product portfolio.

[0067] Sub-step S1023: Decode the feature vector of the financial product portfolio through the decoder layer to obtain the target financial product library.

[0068] By decoding the feature vectors of financial product portfolios through this decoder layer, the target financial product library can be accurately obtained.

[0069] In some embodiments, the market risk value of each financial product group in the target financial product library is determined; based on the market risk value of each financial product group, financial product groups that exceed the preset market risk value of the customer purchasing the financial product are removed from the target financial product library. This preset market risk value is determined based on the user profile. By removing financial product groups that exceed the preset market risk value of the customer purchasing the financial product from the target financial product library, the accuracy of financial product recommendations can be effectively improved, and the interests of customers can be avoided.

[0070] In some embodiments, the market risk value of each financial product group in the target financial product library can be determined by: calculating the average market risk value of each financial product included in each financial product group to obtain the average market risk value of each financial product in each financial product group, and using the average market risk value of each financial product as the market risk value of the financial product group.

[0071] Step S103: Based on the user profile and the target financial product library, a financial product recommendation model is used to recommend financial products to obtain a financial product recommendation list for the customer who has selected financial products.

[0072] In some embodiments, a second sample dataset is obtained, which includes multiple second sample data, including sample user profiles, a sample target financial product library, and an annotated list of financial product recommendations. A preset neural network model is obtained, and a sample data is selected from the second sample dataset as the target sample data. The preset neural network model is used to recommend financial products based on the sample user profiles and the sample target financial product library in the target sample data to obtain a predicted list of financial product recommendations. Based on the annotated list of financial product recommendations and the predicted list of financial product recommendations, it is determined whether the preset neural network model has converged. If the preset neural network model has not converged, the model parameters of the preset neural network model are adjusted, and the step of selecting a sample data from the second sample dataset as the target sample data continues until a converged product recommendation model is obtained.

[0073] It should be noted that the specific training process of this product recommendation model can refer to the training process of the aforementioned product portfolio model, and this application will not elaborate on it further.

[0074] In some embodiments, based on user profiles, a matching score is determined between the customer selecting financial products and each financial product group in the target financial product library; financial product groups with matching scores greater than or equal to a preset matching score are identified as target financial product groups; and the target financial product groups are sorted from high to low according to their matching scores to obtain a financial product recommendation list. The preset matching score can be set according to actual conditions, and this embodiment does not impose specific limitations on it; for example, the preset matching score can be set to 90%. By sorting the target financial product groups from high to low matching scores, a financial product recommendation list can be accurately obtained, greatly improving the efficiency and accuracy of financial product recommendations.

[0075] For example, the system obtains user profiles of customers who purchase financial products and real-time market data of the financial products; it then uses a pre-defined product combination model to combine multiple financial products based on the user profiles and real-time market data to generate a target financial product library. This target financial product library includes financial product 1, financial product 2, financial product 3, financial product group 1 (including financial product 2 and financial product 4), financial product group 2 (including financial product 3, financial product 5 and financial product 7), financial product group 3 (including financial product 3, financial product 6 and financial product 7), and financial product group 4 (including financial product 3, financial product 5, financial product 7 and financial product 8); finally, it uses a pre-defined product recommendation model to recommend financial products based on the user profiles and the target financial product library (financial product 1, financial product 2, financial product 3, financial product group 1, financial product group 2, financial product group 3, and financial product group 4) to obtain a financial product recommendation list for customers who purchase financial products. This financial product recommendation list includes financial product group 3, financial product 2, financial product group 2, and financial product group 1.

[0076] The AI-based financial product recommendation method provided in the above embodiments obtains user profiles of customers selecting financial products and real-time market data of the financial products. This real-time market data includes real-time sales data, yield rates, and market risk values ​​for various types of financial products. Based on the user profiles and the real-time market data, a preset product combination model is used to combine multiple financial products to generate a target financial product library. This target financial product library includes single financial products and groups of multiple financial products. The preset product combination model is obtained by pre-training a preset neural network model based on multiple sample data. Based on the user profiles and the target financial product library, a preset product recommendation model is used to recommend financial products to customers selecting financial products, resulting in a recommended list of financial products for them. This preset product recommendation model is also obtained by pre-training a preset neural network model based on multiple sample data. The product portfolio model in this application combines multiple financial products based on user profiles and real-time market data, which can accurately obtain a target financial product library that matches the market in real time. The product recommendation model recommends financial products from the target financial product library based on user profiles, which can accurately obtain a list of financial product recommendations that match the customers who purchase financial products, greatly improving the efficiency and accuracy of financial product recommendations, and thus greatly improving customers' recognition and satisfaction with financial products.

[0077] Please see Figure 5 , Figure 5 This is a schematic block diagram of a financial product recommendation device provided in an embodiment of this application.

[0078] like Figure 5 As shown, the financial product recommendation device 300 includes an acquisition module 310 and a generation module 320, wherein:

[0079] The acquisition module 310 is used to acquire user profiles of customers who purchase financial products and real-time market data of financial products. The real-time market data includes real-time sales data, yields and market risk values ​​of various types of financial products.

[0080] The generation module 320 is used to combine multiple financial products based on the user profile and the real-time market data using a preset product combination model to generate a target financial product library. The target financial product library includes a single financial product and a financial product group of multiple financial products. The preset product combination model is obtained by training a preset neural network model based on multiple sample data.

[0081] The generation module 320 is further configured to recommend financial products from the target financial product library based on the user profile using a preset product recommendation model, thereby obtaining a list of financial product recommendations for the customer who selects financial products. The preset product recommendation model is obtained by training a preset neural network model based on multiple sample data.

[0082] In some embodiments, the generation module 320 is further configured to:

[0083] The encoder layer performs feature encoding on the user profile and the real-time market data respectively to obtain user feature vectors and real-time market data feature vectors.

[0084] The fully connected layer performs product portfolio prediction on multiple financial products based on the user feature vector and the real-time market data feature vector to obtain the financial product portfolio feature vector.

[0085] The target financial product library is obtained by decoding the feature vector of the financial product portfolio through the decoder layer.

[0086] In some embodiments, the generation module 320 is further configured to:

[0087] Based on the real-time market data feature vector of each financial product, the market value feature vector of each financial product portfolio is determined, wherein the market value feature vector is determined based on the return rate and market risk value of the financial product portfolio.

[0088] Based on the market value feature vector of each financial product portfolio and the user feature vector, the matching degree between the customer who selects the financial products and each financial product portfolio is determined;

[0089] The feature vectors of the financial product portfolio with a matching degree greater than or equal to a preset matching value are fused to obtain the feature vector of the financial product portfolio.

[0090] In some embodiments, the financial product recommendation device 300 is further configured to:

[0091] Determine the market risk value of each financial product group in the target financial product library;

[0092] Based on the market risk value of each financial product group, financial product groups that exceed the preset market risk value of the customers who have selected financial products are removed from the target financial product library. The preset market risk value is determined based on the user profile.

[0093] In some embodiments, the generation module 320 is further configured to:

[0094] Based on the user profile, determine the matching score between the customer who selects financial products and each financial product group in the target financial product library;

[0095] Financial product groups with matching scores greater than or equal to preset matching scores are identified as target financial product groups;

[0096] The target financial product groups are sorted from highest to lowest according to their matching scores to obtain the recommended list of financial products.

[0097] In some embodiments, the acquisition module 310 is further configured to:

[0098] Obtain the asset data, consumption behavior data, and health data of the customers who purchased the financial products;

[0099] Based on the asset data, the consumption behavior data, and the health data, a user profile is constructed for the customer who purchases the financial product.

[0100] In some embodiments, the financial product recommendation device 300 is further configured to:

[0101] Obtain a first sample dataset, which includes multiple first sample data, including sample user profiles, sample real-time market data, and labeled target financial product databases;

[0102] Obtain a preset neural network model, and select a sample data from the first sample dataset as the target sample data;

[0103] Based on the sample user profiles and real-time market data in the target sample data, a preset neural network model is used to combine multiple financial products to generate a predicted target financial product library.

[0104] Based on the predicted target financial product library and the labeled target financial product library, determine whether the preset neural network model has converged;

[0105] If the preset neural network model does not converge, the model parameters of the preset neural network model are adjusted, and the process of selecting a sample data from the first sample dataset as the target sample data continues until a converged product portfolio model is obtained.

[0106] It should be noted that those skilled in the art will understand that, for the sake of convenience and brevity, the specific working process of the aforementioned financial product recommendation device can be referred to the corresponding process in the aforementioned embodiment of the artificial intelligence-based financial product recommendation method, and will not be repeated here.

[0107] Please see Figure 6 , Figure 6This is a schematic block diagram of the structure of a computer device provided in an embodiment of this application.

[0108] like Figure 6 As shown, the computer device 400 includes a processor 402 and a memory 403 connected via a system bus 401, wherein the memory 403 may include a storage medium and internal memory.

[0109] The storage medium may store a computer program. This computer program includes program instructions that, when executed, cause the processor to perform any artificial intelligence-based financial product recommendation method.

[0110] Processor 402 provides computing and control capabilities to support the operation of the entire computer device.

[0111] Internal memory provides an environment for the execution of computer programs stored in the storage medium. When these computer programs are executed by the processor, the processor can perform any artificial intelligence-based financial product recommendation method.

[0112] Those skilled in the art will understand that Figure 6 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0113] It should be understood that processor 402 can be a Central Processing Unit (CPU), but it can also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among these, the general-purpose processor can be a microprocessor or any conventional processor.

[0114] In one embodiment, the processor 402 is configured to run a computer program stored in a memory to perform the following steps:

[0115] Acquire user profiles of customers who purchase financial products and real-time market data of financial products, including real-time sales data, yields and market risk values ​​of various types of financial products.

[0116] Based on the user profile and the real-time market data, multiple financial products are combined using a preset product combination model to generate a target financial product library. The target financial product library includes single financial products and financial product groups of multiple financial products. The preset product combination model is obtained by training a preset neural network model based on multiple sample data.

[0117] Based on the user profile and the target financial product library, a preset product recommendation model is used to recommend financial products, thereby obtaining a list of recommended financial products for the customer who has selected financial products. The preset product recommendation model is obtained by training a preset neural network model based on multiple sample data.

[0118] In one embodiment, the preset product portfolio model includes an encoder layer, a fully connected layer, and a decoder layer; the processor 402, when implementing the process of combining multiple financial products based on the user profile and the real-time market data using the preset product portfolio model to generate a target financial product library, is used to:

[0119] The encoder layer performs feature encoding on the user profile and the real-time market data respectively to obtain user feature vectors and real-time market data feature vectors.

[0120] The fully connected layer performs product portfolio prediction on multiple financial products based on the user feature vector and the real-time market data feature vector to obtain the financial product portfolio feature vector.

[0121] The target financial product library is obtained by decoding the feature vector of the financial product portfolio through the decoder layer.

[0122] In one embodiment, when implementing the process of predicting a financial product portfolio feature vector by performing product portfolio prediction on multiple financial products through the fully connected layer based on the user feature vector and the real-time market data feature vector, the processor 402 is used to:

[0123] Based on the real-time market data feature vector of each financial product, the market value feature vector of each financial product portfolio is determined, wherein the market value feature vector is determined based on the return rate and market risk value of the financial product portfolio.

[0124] Based on the market value feature vector of each financial product portfolio and the user feature vector, the matching degree between the customer who selects the financial products and each financial product portfolio is determined;

[0125] The feature vectors of the financial product portfolio with a matching degree greater than or equal to a preset matching value are fused to obtain the feature vector of the financial product portfolio.

[0126] In one embodiment, after implementing the process of combining multiple financial products based on the user profile and the real-time market data using a preset product portfolio model to generate a target financial product library, the processor 402 is further configured to implement:

[0127] Determine the market risk value of each financial product group in the target financial product library;

[0128] Based on the market risk value of each financial product group, financial product groups that exceed the preset market risk value of the customers who have selected financial products are removed from the target financial product library. The preset market risk value is determined based on the user profile.

[0129] In one embodiment, when the processor 402 performs the step of recommending financial products based on the user profile and the target financial product library using a preset product recommendation model to obtain a list of recommended financial products for the customer who has selected financial products, it is configured to:

[0130] Based on the user profile, determine the matching score between the customer who selects financial products and each financial product group in the target financial product library;

[0131] Financial product groups with matching scores greater than or equal to preset matching scores are identified as target financial product groups;

[0132] The target financial product groups are sorted from highest to lowest according to their matching scores to obtain the recommended list of financial products.

[0133] In one embodiment, when the processor 402 acquires the user profile of a customer who has selected financial products, it is configured to:

[0134] Obtain the asset data, consumption behavior data, and health data of the customers who purchased the financial products;

[0135] Based on the asset data, the consumption behavior data, and the health data, a user profile is constructed for the customer who purchases the financial product.

[0136] In one embodiment, the processor 402 is further configured to implement:

[0137] Obtain a first sample dataset, which includes multiple first sample data, including sample user profiles, sample real-time market data, and labeled target financial product databases;

[0138] Obtain a preset neural network model, and select a sample data from the first sample dataset as the target sample data;

[0139] Based on the sample user profiles and real-time market data in the target sample data, a preset neural network model is used to combine multiple financial products to generate a predicted target financial product library.

[0140] Based on the predicted target financial product library and the labeled target financial product library, determine whether the preset neural network model has converged;

[0141] If the preset neural network model fails to converge, the model parameters of the preset neural network model are adjusted, and the process of selecting a sample data from the first sample dataset as the target sample data continues until a converged product portfolio model is obtained.

[0142] It should be noted that those skilled in the art will understand that, for the sake of convenience and brevity, the specific working process of the computer equipment described above can be referred to the corresponding process in the aforementioned embodiment of the AI-based financial product recommendation method, and will not be repeated here.

[0143] This application also provides a computer-readable storage medium storing a computer program, the computer program including program instructions, and the method implemented when the program instructions are executed can be referred to the various embodiments of the artificial intelligence-based financial product recommendation method of this application.

[0144] The computer-readable storage medium can be an internal storage unit of the computer device described in the foregoing embodiments, such as a hard disk or memory of the computer device. The computer-readable storage medium can be non-volatile or volatile. Alternatively, the computer-readable storage medium can be an external storage device of the computer device, such as a plug-in hard disk, Smart Media Card (SMC), Secure Digital (SD) card, or Flash Card equipped on the computer device.

[0145] Furthermore, the computer-readable storage medium may primarily include a program storage area and a data storage area, wherein the program storage area may store the operating system, at least one application required for a function, etc.; and the data storage area may store data created based on the use of blockchain nodes, etc.

[0146] The blockchain referred to in this invention is a novel application model of computer technologies such as distributed data storage, peer-to-peer transmission, consensus mechanisms, and encryption algorithms. Essentially, a blockchain is a decentralized database, a chain of data blocks linked together using cryptographic methods. Each data block contains information about a batch of network transactions, used to verify the validity of the information (anti-counterfeiting) and generate the next block. A blockchain can include an underlying blockchain platform, a platform product service layer, and an application service layer.

[0147] It should be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the scope of the application. As used in this specification, unless the context clearly indicates otherwise, the singular forms “a,” “an,” and “the” are intended to include the plural forms.

[0148] It should also be understood that the term "and / or" as used in this specification refers to any combination and all possible combinations of one or more of the associated listed items, and includes such combinations. It should be noted that, herein, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or system. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or system that includes that element.

[0149] The sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments. The above descriptions are merely specific implementations of this application, but the protection scope of this application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and these modifications or substitutions should all be covered within the protection scope of this application.

Claims

1. A method for recommending financial products based on artificial intelligence, characterized in that, include: Acquire user profiles of customers who purchase financial products and real-time market data of financial products, including real-time sales data, yields and market risk values ​​of various types of financial products. Based on the user profile and the real-time market data, multiple financial products are combined using a preset product combination model to generate a target financial product library. The target financial product library includes single financial products and financial product groups of multiple financial products. The preset product combination model is obtained by training a preset neural network model based on multiple sample data. Based on the user profile and the target financial product library, a preset product recommendation model is used to recommend financial products, thereby obtaining a list of recommended financial products for the customer who has selected financial products. The preset product recommendation model is obtained by training a preset neural network model based on multiple sample data.

2. The artificial intelligence-based financial product recommendation method as described in claim 1, characterized in that, The preset product portfolio model includes an encoder layer, a fully connected layer, and a decoder layer; the process of combining multiple financial products based on the user profile and the real-time market data using the preset product portfolio model to generate a target financial product library includes: The encoder layer performs feature encoding on the user profile and the real-time market data respectively to obtain user feature vectors and real-time market data feature vectors. The fully connected layer performs product portfolio prediction on multiple financial products based on the user feature vector and the real-time market data feature vector to obtain the financial product portfolio feature vector. The target financial product library is obtained by decoding the feature vector of the financial product portfolio through the decoder layer.

3. The artificial intelligence-based financial product recommendation method as described in claim 2, characterized in that, The step of performing product portfolio prediction on multiple financial products through the fully connected layer based on the user feature vector and the real-time market data feature vector to obtain a financial product portfolio feature vector includes: Based on the real-time market data feature vector of each financial product, the market value feature vector of each financial product portfolio is determined, wherein the market value feature vector is determined based on the return rate and market risk value of the financial product portfolio. Based on the market value feature vector of each financial product portfolio and the user feature vector, the matching degree between the customer who selects the financial products and each financial product portfolio is determined; The feature vectors of the financial product portfolio with a matching degree greater than or equal to a preset matching value are fused to obtain the feature vector of the financial product portfolio.

4. The artificial intelligence-based financial product recommendation method as described in claim 1, characterized in that, After generating a target financial product library by combining multiple financial products based on the user profile and the real-time market data using a preset product portfolio model, the process further includes: Determine the market risk value of each financial product group in the target financial product library; Based on the market risk value of each financial product group, financial product groups that exceed the preset market risk value of the customers who have selected financial products are removed from the target financial product library. The preset market risk value is determined based on the user profile.

5. The artificial intelligence-based financial product recommendation method as described in claim 1, characterized in that, The step involves using a preset product recommendation model to recommend financial products based on the user profile and the target financial product database, resulting in a financial product recommendation list for the customer who intends to purchase financial products. This list includes: Based on the user profile, determine the matching score between the customer who selects financial products and each financial product group in the target financial product library; Financial product groups with matching scores greater than or equal to preset matching scores are identified as target financial product groups; The target financial product groups are sorted from highest to lowest according to their matching scores to obtain the recommended list of financial products.

6. The method for recommending financial products based on artificial intelligence as described in claim 1, characterized in that, The process of obtaining user profiles of customers who purchase financial products includes: Obtain the asset data, consumption behavior data, and health data of the customers who purchased the financial products; Based on the asset data, the consumption behavior data, and the health data, a user profile is constructed for the customer who purchases the financial product.

7. The method for recommending financial products based on artificial intelligence as described in any one of claims 1-6, characterized in that, The method further includes: Obtain a first sample dataset, which includes multiple first sample data, including sample user profiles, sample real-time market data, and labeled target financial product databases; Obtain a preset neural network model, and select a sample data from the first sample dataset as the target sample data; Based on the sample user profiles and real-time market data in the target sample data, a preset neural network model is used to combine multiple financial products to generate a predicted target financial product library. Based on the predicted target financial product library and the labeled target financial product library, determine whether the preset neural network model has converged; If the preset neural network model does not converge, the model parameters of the preset neural network model are adjusted, and the process of selecting a sample data from the first sample dataset as the target sample data continues until a converged product portfolio model is obtained.

8. A financial product recommendation device, characterized in that, The financial product recommendation device includes an acquisition module and a generation module, wherein: The acquisition module is used to acquire user profiles of customers who purchase financial products and real-time market data of financial products. The real-time market data includes real-time sales data, yields, and market risk values ​​of various types of financial products. The generation module is used to combine multiple financial products based on the user profile and the real-time market data using a preset product combination model to generate a target financial product library. The target financial product library includes single financial products and financial product groups of multiple financial products. The preset product combination model is obtained by training a preset neural network model based on multiple sample data in advance. The generation module is further configured to recommend financial products from the target financial product library based on the user profile using a preset product recommendation model, thereby obtaining a list of financial product recommendations for the customer who selects financial products. The preset product recommendation model is obtained by training a preset neural network model based on multiple sample data.

9. A computer device, characterized in that, The computer device includes a processor, a memory, and a computer program stored in the memory and executable by the processor, wherein when the computer program is executed by the processor, it implements the steps of the artificial intelligence-based financial product recommendation method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, wherein when the computer program is executed by a processor, it implements the steps of the artificial intelligence-based financial product recommendation method as described in any one of claims 1 to 7.

Images