CN-121998729-A - Recommendation method and related device for financial products

CN121998729ACN 121998729 ACN121998729 ACN 121998729ACN-121998729-A

Abstract

The application discloses a recommendation method and a related device of financial products, wherein a target financial product is determined from a plurality of candidate financial products according to financial preference characteristics of an object, product characteristics respectively corresponding to the candidate financial products and historical income characteristics respectively corresponding to the candidate financial products in a historical period. The interest preference of the object is considered, whether the financial product accords with the interest preference is considered, the income performance of the financial product is considered, the product characteristics of the target financial product and the financial preference characteristics which are determined correspondingly meet the preference screening condition, and the predicted income characteristics of the target financial product in the future period meet the income screening condition, namely the target financial product is more likely to have good income performance in the future period. Finally, recommending the target financial product to the object. The financial product profit performance is considered, so that the problem that the financial product profit performance is poor and the financial product profit performance is not in line with the user expectation is avoided, and the recommendation accuracy is improved.

Inventors

LIN YUE

Assignees

腾讯科技（深圳）有限公司

Dates

Publication Date: 20260508
Application Date: 20241107

Claims (17)

1. A method of recommending a financial product, the method comprising: Acquiring financial preference characteristics corresponding to an object, and acquiring product characteristics respectively corresponding to a plurality of candidate financial products and historical income characteristics respectively corresponding to a plurality of candidate financial products in a historical period; Determining a target financial product from a plurality of candidate financial products according to the financial preference characteristics, product characteristics respectively corresponding to a plurality of candidate financial products and historical revenue characteristics respectively corresponding to a plurality of candidate financial products, wherein a preference screening condition is met between the product characteristics of the target financial product and the financial preference characteristics, and a predicted revenue characteristic of the target financial product in a future period meets a revenue screening condition, and the predicted revenue characteristic corresponding to the target financial product is determined according to the historical revenue characteristic corresponding to the target financial product; recommending the target financial product to the subject.
2. The method of claim 1, wherein the obtaining the financial preference feature corresponding to the object comprises: Acquiring financial transaction data corresponding to the object and interaction data of the object aiming at financial information; and determining the financial preference characteristics according to the financial transaction data and the interaction data.
3. The method according to claim 2, wherein the method further comprises: Obtaining market state characteristics corresponding to a financial market; Determining a first weight corresponding to the financial transaction data and a second weight corresponding to the interaction data according to the market state characteristics, wherein different market state characteristics correspond to different weight combinations, and the weight combinations comprise the first weight and the second weight; Said determining said financial preference characteristics from said financial transaction data and said interaction data comprising: the financial preference feature is determined based on the financial transaction data, the first weight, the interaction data, and the second weight.
4. A method according to claim 3, wherein determining a first weight corresponding to the financial transaction data and a second weight corresponding to the interaction data based on the market state characteristics comprises: determining a sub-model matched with the market state characteristics from a plurality of sub-models included in the weight model as a target sub-model according to the market state characteristics, wherein the sub-models matched with different market state characteristics are different; According to the financial transaction data and the interaction data, determining a first weight corresponding to the financial transaction data and a second weight corresponding to the interaction data through the target sub-model; Wherein the target submodel is determined by: Obtaining a first training sample with a first sample tag, the first training sample comprising first sample financial transaction data and first sample interaction data, the first sample tag being used to indicate a first sample weight corresponding to the first sample financial transaction data under the market state feature and a second sample weight corresponding to the first sample interaction data under the market state feature, different market state features corresponding to different sample weight combinations, the sample weight combinations comprising the first sample weight and the second sample weight; according to the first training sample, determining a first prediction weight corresponding to the first sample melting transaction data and a second prediction weight corresponding to the first sample interaction data through an initial sub-model; And carrying out model training on the initial sub-model according to the difference among the first prediction weight, the second prediction weight and the first sample label to obtain the target sub-model.
5. The method of claim 3, wherein the obtaining market state features corresponding to the financial market comprises: Obtaining market fluctuation parameters and market trend parameters corresponding to the financial market, wherein the market fluctuation parameters are used for indicating the fluctuation degree of the financial market in the historical period, and the market trend parameters are used for indicating the trend direction of the financial market in the future period; and determining the market state characteristics according to the market fluctuation parameters and the market trend parameters.
6. The method of claim 1, wherein the obtaining the financial preference feature corresponding to the object comprises: Acquiring a plurality of interaction data of the object aiming at financial information; And determining financial preference characteristics corresponding to the objects according to the interaction data.
7. The method of claim 6, wherein if the plurality of interaction data are multimodal data, the determining the financial preference characteristics corresponding to the object according to the plurality of interaction data comprises: according to the interaction data, an interaction graph structure is constructed through a graph neural network, nodes in the interaction graph structure are used for marking the interaction data, and if the interaction data marked by a first node and the interaction data marked by a second node are data in different modes and edges are arranged between the first node and the second node, the edges are used for indicating the description relationship between the interaction data marked by the first node and the interaction data marked by the second node; determining target interaction characteristics through the graph neural network according to the interaction graph structure; and determining the financial preference characteristic according to the target interaction characteristic.
8. The method of claim 6, wherein determining the financial preference characteristics corresponding to the object from the plurality of interaction data comprises: Respectively extracting features of the interaction data to obtain a plurality of interaction features; Determining fusion weights corresponding to the interaction characteristics according to the occurrence time of the interaction data and the interaction type of the interaction data; and determining the financial preference characteristics according to the interaction characteristics and the fusion weights respectively corresponding to the interaction characteristics.
9. The method of claim 6, wherein determining the financial preference characteristics corresponding to the object from the plurality of interaction data comprises: clustering the interaction data to obtain a plurality of interaction clusters, wherein the interaction clusters are used for indicating interaction behavior characteristics of the object aiming at the financial information of the target type, and different interaction clusters correspond to different target types; And determining the financial preference characteristics of the object according to a plurality of interaction clusters.
10. The method of any one of claims 1-9, wherein if the candidate financial product is of a target financial type, the method further comprises: acquiring the corresponding income state characteristics of the target financial type financial products in the financial market as the income state characteristics corresponding to the candidate financial products; The determining a target financial product from the plurality of candidate financial products according to the financial preference characteristics, the product characteristics respectively corresponding to the plurality of candidate financial products, and the historical revenue characteristics respectively corresponding to the plurality of candidate financial products comprises: And determining the target financial product from the candidate financial products according to the financial preference characteristics, the product characteristics corresponding to the candidate financial products respectively, the historical profit characteristics corresponding to the candidate financial products respectively and the profit state characteristics corresponding to the candidate financial products respectively, wherein the predicted profit characteristics corresponding to the target financial product are determined according to the historical profit characteristics corresponding to the target financial product and the profit state characteristics corresponding to the target financial product.
11. The method according to any one of claims 1-9, further comprising: acquiring historical transaction data corresponding to the candidate financial products in the historical period and product risk data corresponding to the candidate financial products; The determining a target financial product from the plurality of candidate financial products according to the financial preference characteristics, the product characteristics respectively corresponding to the plurality of candidate financial products, and the historical revenue characteristics respectively corresponding to the plurality of candidate financial products comprises: And determining the target financial product from the candidate financial products according to the financial preference characteristics, the product characteristics corresponding to the candidate financial products respectively, the historical revenue characteristics corresponding to the candidate financial products respectively, the historical achievement data corresponding to the candidate financial products respectively and the product risk data corresponding to the candidate financial products respectively, wherein the predicted revenue characteristics corresponding to the target financial product are determined according to the historical revenue characteristics corresponding to the target financial product, the historical achievement data corresponding to the target financial product and the product risk data corresponding to the target financial product.
12. The method of any of claims 1-9, wherein the determining a target financial product from a plurality of candidate financial products based on the financial preference characteristics, product characteristics respectively corresponding to a plurality of candidate financial products, and historical revenue characteristics respectively corresponding to a plurality of candidate financial products comprises: Determining a prediction matching degree corresponding to the candidate financial product through a prediction model according to the financial preference characteristic, the product characteristic corresponding to the candidate financial product respectively and the historical revenue characteristic corresponding to the candidate financial product respectively, wherein the prediction matching degree is used for indicating whether the product characteristic of the candidate financial product and the financial preference characteristic meet the preference screening condition or not and whether the predicted revenue characteristic of the candidate financial product in the future period meets the revenue screening condition or not; Screening candidate financial products with the predicted matching degree larger than a matching degree threshold value as the target financial products according to the predicted matching degrees respectively corresponding to the candidate financial products; Wherein the predictive model is determined by: obtaining a second training sample having a second sample tag, the second training sample comprising sample financial preference characteristics of a sample object, sample product characteristics of a sample financial product, and sample historical revenue characteristics of the sample financial product, the second sample tag being used to indicate a degree of matching between the sample financial product and the sample object; according to the second training sample, determining the prediction matching degree corresponding to the sample financial product through an initial model; and carrying out model training on the initial model according to the difference between the prediction matching degree corresponding to the sample financial product and the second sample label to obtain the prediction model.
13. The method of any of claims 1-9, wherein the determining a target financial product from a plurality of candidate financial products based on the financial preference characteristics, product characteristics respectively corresponding to a plurality of candidate financial products, and historical revenue characteristics respectively corresponding to a plurality of candidate financial products comprises: Screening candidate financial products meeting the preference screening conditions between the product characteristics and the financial preference characteristics as pending financial products according to the financial preference characteristics and product characteristics respectively corresponding to a plurality of candidate financial products; and screening the pending financial products with the predicted revenue characteristics meeting the revenue screening conditions as the target financial products according to the historical revenue characteristics corresponding to the pending financial products.
14. A recommendation device for a financial product, the device comprising an acquisition unit, a determination unit and a recommendation unit: The acquisition unit is used for acquiring financial preference characteristics corresponding to the object, product characteristics corresponding to a plurality of candidate financial products respectively and historical income characteristics corresponding to the candidate financial products respectively in a historical period; The determining unit is configured to determine a target financial product from the plurality of candidate financial products according to the financial preference characteristics, product characteristics corresponding to the plurality of candidate financial products respectively, and historical revenue characteristics corresponding to the plurality of candidate financial products respectively, where a preference screening condition is satisfied between the product characteristics of the target financial product and the financial preference characteristics, and a predicted revenue characteristic of the target financial product in a future period satisfies a revenue screening condition, and the predicted revenue characteristic corresponding to the target financial product is determined according to the historical revenue characteristic corresponding to the target financial product; the recommending unit is used for recommending the target financial product to the object.
15. A computer device, the computer device comprising a processor and a memory: The memory is used for storing a computer program and transmitting the computer program to the processor; The processor is configured to perform the method of any of claims 1-13 according to instructions in the computer program.
16. A computer readable storage medium for storing a computer program which, when executed by a computer device, causes the computer device to perform the method of any of claims 1-13.
17. A computer program product comprising a computer program which, when run on a computer device, causes the computer device to perform the method of any of claims 1-13.

Description

Recommendation method and related device for financial products Technical Field The application relates to the technical field of data processing, in particular to a recommendation method and a related device for financial products. Background With the rapid development of internet technology, users can know on-line financial products related to the financial field, such as credit products, financial products, etc. In practical applications, a user may initiate a credit application, conduct financial management, or the like through an application provided by a institution such as a bank, or a financial website, or the like. Accordingly, how to recommend the proper financial products to the user is beneficial to the user to quickly find the proper financial products or the financial products required by the user, and meanwhile, the development of financial institutions (such as banking institutions) and the like can be promoted. In the related art, a financial product is recommended to a user based on interest preferences of the user and a recommendation rule defined in advance. The recommendation rules may be, for example, what financial products are recommended to a user group of what interest preferences. However, the recommendation method adopted in the related art has a problem that it is difficult to recommend the financial product which meets the user's desire to the user, that is, the recommendation accuracy is poor. Disclosure of Invention In order to solve the technical problems, the application provides a recommendation method and a related device for financial products, which can avoid the problem that the financial products with poor profit performance are recommended to users and are not in line with the expectations of the users, so that the recommendation accuracy can be improved. The embodiment of the application discloses the following technical scheme: In one aspect, an embodiment of the present application provides a recommendation method for a financial product, the method including: Acquiring financial preference characteristics corresponding to an object, and acquiring product characteristics respectively corresponding to a plurality of candidate financial products and historical income characteristics respectively corresponding to a plurality of candidate financial products in a historical period; Determining a target financial product from a plurality of candidate financial products according to the financial preference characteristics, product characteristics respectively corresponding to a plurality of candidate financial products and historical revenue characteristics respectively corresponding to a plurality of candidate financial products, wherein a preference screening condition is met between the product characteristics of the target financial product and the financial preference characteristics, and a predicted revenue characteristic of the target financial product in a future period meets a revenue screening condition, and the predicted revenue characteristic corresponding to the target financial product is determined according to the historical revenue characteristic corresponding to the target financial product; recommending the target financial product to the subject. In still another aspect, an embodiment of the present application provides a recommendation apparatus for a financial product, including an acquisition unit, a determination unit, and a recommendation unit: The acquisition unit is used for acquiring financial preference characteristics corresponding to the object, product characteristics corresponding to a plurality of candidate financial products respectively and historical income characteristics corresponding to the candidate financial products respectively in a historical period; The determining unit is configured to determine a target financial product from the plurality of candidate financial products according to the financial preference characteristics, product characteristics corresponding to the plurality of candidate financial products respectively, and historical revenue characteristics corresponding to the plurality of candidate financial products respectively, where a preference screening condition is satisfied between the product characteristics of the target financial product and the financial preference characteristics, and a predicted revenue characteristic of the target financial product in a future period satisfies a revenue screening condition, and the predicted revenue characteristic corresponding to the target financial product is determined according to the historical revenue characteristic corresponding to the target financial product; the recommending unit is used for recommending the target financial product to the object. In a possible implementation manner, the obtaining unit is further configured to: Acquiring financial transaction data corresponding to the object and interaction data of the object aiming at financial information