CN117952719A - Recommendation method and device for financial products, storage medium and electronic equipment - Google Patents

Recommendation method and device for financial products, storage medium and electronic equipment Download PDF

Info

Publication number
CN117952719A
CN117952719A CN202410177654.5A CN202410177654A CN117952719A CN 117952719 A CN117952719 A CN 117952719A CN 202410177654 A CN202410177654 A CN 202410177654A CN 117952719 A CN117952719 A CN 117952719A
Authority
CN
China
Prior art keywords
feature vector
target
initial
financial products
financial
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202410177654.5A
Other languages
Chinese (zh)
Inventor
王若辰
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Industrial and Commercial Bank of China Ltd ICBC
Original Assignee
Industrial and Commercial Bank of China Ltd ICBC
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Industrial and Commercial Bank of China Ltd ICBC filed Critical Industrial and Commercial Bank of China Ltd ICBC
Priority to CN202410177654.5A priority Critical patent/CN117952719A/en
Publication of CN117952719A publication Critical patent/CN117952719A/en
Pending legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce
    • G06Q30/06—Buying, selling or leasing transactions
    • G06Q30/0601—Electronic shopping [e-shopping]
    • G06Q30/0631—Item recommendations
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
    • G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/25—Fusion techniques
    • G06F18/253—Fusion techniques of extracted features
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/042—Knowledge-based neural networks; Logical representations of neural networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/045—Combinations of networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
    • G06Q40/06—Asset management; Financial planning or analysis
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Business, Economics & Management (AREA)
  • General Physics & Mathematics (AREA)
  • Artificial Intelligence (AREA)
  • Accounting & Taxation (AREA)
  • Evolutionary Computation (AREA)
  • General Engineering & Computer Science (AREA)
  • Finance (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Development Economics (AREA)
  • Molecular Biology (AREA)
  • Economics (AREA)
  • Computational Linguistics (AREA)
  • General Health & Medical Sciences (AREA)
  • Biomedical Technology (AREA)
  • Computing Systems (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Health & Medical Sciences (AREA)
  • Evolutionary Biology (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • General Business, Economics & Management (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Strategic Management (AREA)
  • Marketing (AREA)
  • Biophysics (AREA)
  • Operations Research (AREA)
  • Human Resources & Organizations (AREA)
  • Game Theory and Decision Science (AREA)
  • Technology Law (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

The application discloses a recommendation method and device of financial products, a storage medium and electronic equipment, and relates to the technical fields of artificial intelligence, financial science and technology and other related technical fields, wherein the method comprises the following steps: constructing a target map according to the historical behavior data and all financial products in the target database; processing the target map through a graph neural network model to obtain a feature vector set; the first feature vector and the second feature vector are spliced through the graph neural network model to obtain a target feature vector, and the preference score of the first object to each financial product in the target database is obtained according to the target feature vector; and screening from the financial products in the target database according to the preference scores to determine target financial products to be recommended. The application solves the problem that the accuracy of recommending the financial products is lower because the financial products purchased by the user are recommended to the user in the related technology.

Description

Recommendation method and device for financial products, storage medium and electronic equipment
Technical Field
The application relates to the technical field of artificial intelligence, the technical field of financial science and technology and other related technical fields, in particular to a recommendation method and device of financial products, a storage medium and electronic equipment.
Background
With the rapid development of technology, data and information all show a trend of a great increase, under the background that the user becomes difficult to acquire information, the phenomenon of information overload is serious, and how to acquire proper information from massive data is a great difficulty, so that a search engine and a recommendation system are common methods for solving the phenomenon, and the possible wanted information is provided for the user mainly through the preference information and the information characteristics of the user. With the rapid development of the fund industry, the development of the internet injects new vitality into the fund industry, and a few users consider purchasing financial products, but the conventional algorithm recommends financial products to users based on the financial products purchased by the users, and the mode often has the problem of low accuracy of recommending the financial products.
Aiming at the problem that the accuracy of recommending financial products to users is low because the financial products purchased by the users are recommended to the users in the related art, no effective solution is proposed at present.
Disclosure of Invention
The application mainly aims to provide a recommendation method and device for financial products, a storage medium and electronic equipment, and aims to solve the problem that in the related art, the accuracy of recommending financial products is low because financial products purchased by users are recommended to users.
In order to achieve the above object, according to one aspect of the present application, there is provided a recommendation method of a financial product. The method comprises the following steps: acquiring historical behavior data of a first object on financial products in a target database, and constructing a target map according to the historical behavior data and all the financial products in the target database; processing the target map through a graph neural network model to obtain a feature vector set, wherein the feature vector set at least comprises: a first feature vector corresponding to the first object and a second feature vector corresponding to each financial product in the target database; the first feature vector and the second feature vector are spliced through the graph neural network model to obtain a target feature vector, and the preference score of the first object to each financial product in a target database is obtained according to the target feature vector; and screening from the financial products in the target database according to the preference scores to determine target financial products to be recommended.
Further, constructing a target map according to the historical behavior data and all financial products in the target database includes: determining a plurality of second objects corresponding to each financial product in the target database, wherein the second objects are objects having interaction behaviors with the financial products in the target database; acquiring first scoring data of the second object on the financial products in the target database according to the interaction behavior data between the second object and the financial products in the target database; acquiring second scoring data of the first object on financial products corresponding to the historical behavior data according to the historical behavior data; and constructing the target map according to the historical behavior data, the first scoring data and the second scoring data.
Further, processing the target map through a graph neural network model to obtain a feature vector set includes: extracting features of the target map through the graph neural network model to obtain a first initial feature vector corresponding to the first object, a second initial feature vector corresponding to each financial product in the target database, a third initial feature vector and a fourth initial feature vector, wherein the third initial feature vector is a feature vector corresponding to the first scoring data, and the fourth initial feature vector is a feature vector corresponding to the second scoring data; determining the first feature vector according to the second initial feature vector and the third initial feature vector; determining the second feature vector according to the first initial feature vector and the fourth initial feature vector; and determining the feature vector set according to the first feature vector and the second feature vector.
Further, determining the first feature vector from the second initial feature vector and the third initial feature vector includes: calculating according to the second initial feature vector and the third initial feature vector to obtain a first interaction feature vector of the first object on the financial product corresponding to the historical behavior data; calculating according to the interaction feature vector and the first initial feature vector to obtain target scoring data of the first object on the financial product corresponding to the historical behavior data; and determining the first feature vector according to the target scoring data.
Further, determining the first feature vector according to the target scoring data includes: calculating according to the target scoring data to obtain a target proportion value of the financial product corresponding to the historical behavior data; determining a plurality of initial financial products from the financial products corresponding to the historical behavior data according to the target specific gravity value; and carrying out aggregation processing on the first interaction feature vectors corresponding to the plurality of initial financial products to obtain the first feature vectors.
Further, performing aggregation processing on the first interaction feature vectors corresponding to the plurality of initial financial products, and obtaining the first feature vectors includes: acquiring a target weight value corresponding to each initial financial product through a multi-head attention mechanism, wherein the target weight value is used for representing the contribution degree of the first feature vector; and carrying out aggregation processing according to the target weight value corresponding to each initial financial product and the first interaction feature vector to obtain the first feature vector.
Further, determining the second feature vector from the first initial feature vector and the fourth initial feature vector comprises: calculating according to the first initial feature vector and the fourth initial feature vector to obtain a second interaction feature vector of the second object on each financial product in the target database; and determining the second feature vector according to the second interaction feature vector.
In order to achieve the above object, according to another aspect of the present application, there is provided a recommendation device for a financial product. The device comprises: the acquisition unit is used for acquiring historical behavior data of the first object on the financial products in the target database and constructing a target map according to the historical behavior data and all the financial products in the target database; the processing unit is used for processing the target map through a graph neural network model to obtain a feature vector set, wherein the feature vector set at least comprises: a first feature vector corresponding to the first object and a second feature vector corresponding to each financial product in the target database; the splicing unit is used for carrying out splicing processing on the first feature vector and the second feature vector through the graph neural network model to obtain a target feature vector, and obtaining the preference score of the first object on each financial product in a target database according to the target feature vector; and the screening unit is used for screening from the financial products in the target database according to the preference scores and determining target financial products to be recommended.
Further, the acquisition unit includes: a determining subunit, configured to determine a plurality of second objects corresponding to each financial product in the target database, where the second objects are objects that have interaction behaviors with the financial products in the target database; the first obtaining subunit is used for obtaining first scoring data of the second object on the financial products in the target database according to the interaction behavior data between the second object and the financial products in the target database; the second obtaining subunit is used for obtaining second scoring data of the financial products corresponding to the historical behavior data by the first object according to the historical behavior data; the construction unit is used for constructing the target map according to the historical behavior data, the first scoring data and the second scoring data.
Further, the processing unit includes: the extraction subunit is used for extracting the characteristics of the target map through the graph neural network model to obtain a first initial characteristic vector corresponding to the first object, a second initial characteristic vector, a third initial characteristic vector and a fourth initial characteristic vector corresponding to each financial product in the target database, wherein the third initial characteristic vector is a characteristic vector corresponding to the first scoring data, and the fourth initial characteristic vector is a characteristic vector corresponding to the second scoring data; a first determining subunit, configured to determine the first feature vector according to the second initial feature vector and the third initial feature vector; a second determining subunit, configured to determine the second feature vector according to the first initial feature vector and the fourth initial feature vector; and the third determination subunit is used for determining the feature vector set according to the first feature vector and the second feature vector.
Further, the first determining subunit includes: the first computing module is used for computing according to the second initial feature vector and the third initial feature vector to obtain a first interaction feature vector of the first object on the financial product corresponding to the historical behavior data; the second calculation module is used for calculating according to the interaction feature vector and the first initial feature vector to obtain target scoring data of the first object on the financial product corresponding to the historical behavior data; and the first determining module is used for determining the first feature vector according to the target scoring data.
Further, the first determining module includes: the calculation sub-module is used for calculating according to the target scoring data to obtain a target specific gravity value of the financial product corresponding to the historical behavior data; the first determining submodule is used for determining a plurality of initial financial products from the financial products corresponding to the historical behavior data according to the target specific gravity value; and the aggregation sub-module is used for conducting aggregation processing on the first interaction feature vectors corresponding to the plurality of initial financial products to obtain the first feature vectors.
Further, the aggregation sub-module includes: the acquisition sub-module is used for acquiring a target weight value corresponding to each initial financial product through a multi-head attention mechanism, wherein the target weight value is used for representing the contribution degree of the first feature vector; and the aggregation sub-module is used for carrying out aggregation processing according to the target weight value corresponding to each initial financial product and the first interaction feature vector to obtain the first feature vector.
Further, the second determining subunit includes: the third calculation module is used for calculating according to the first initial feature vector and the fourth initial feature vector to obtain a second interaction feature vector of the second object on each financial product in the target database; and the second determining module is used for determining the second characteristic vector according to the second interaction characteristic vector.
According to the application, the following steps are adopted: acquiring historical behavior data of a first object on financial products in a target database, and constructing a target map according to the historical behavior data and all the financial products in the target database; processing the target map through the graph neural network model to obtain a feature vector set, wherein the feature vector set at least comprises: a first feature vector corresponding to the first object and a second feature vector corresponding to each financial product in the target database; the first feature vector and the second feature vector are spliced through the graph neural network model to obtain a target feature vector, and the preference score of the first object to each financial product in the target database is obtained according to the target feature vector; the target financial products to be recommended are determined by screening from the financial products in the target database according to the preference scores, and the problem that the accuracy of the recommended financial products is low due to the fact that the financial products purchased by the user are recommended to the user in the related technology is solved. According to the technical scheme, historical behavior data of a first object on financial products in a target database are collected, a target map is established according to the historical behavior data and all the financial products in the target database, the association relation between the first object and the products is described through the target map, then feature extraction is carried out on the target map through a graph neural network model, a first feature vector corresponding to the first object and a second feature vector corresponding to the financial products are obtained, further preference scores of the first object on each financial product in the target database are obtained through the first feature vector and the second feature vector, finally the financial products to be recommended are screened from the financial products in the target database according to the preference scores, the association degree between users and the products can be captured more accurately through the target map, and the effect of improving the accuracy of recommending the financial products is achieved.
Drawings
The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this specification, illustrate embodiments of the application and together with the description serve to explain the application. In the drawings:
FIG. 1 is a flow chart of a recommendation method for financial products provided according to an embodiment of the present application;
FIG. 2 is a flow chart of an alternative method of recommending a financial product provided in accordance with an embodiment of the present application;
FIG. 3 is a schematic diagram of a recommendation device for financial products provided according to an embodiment of the present application;
Fig. 4 is a schematic diagram of an electronic device according to an embodiment of the present application.
Detailed Description
It should be noted that, without conflict, the embodiments of the present application and features of the embodiments may be combined with each other. The application will be described in detail below with reference to the drawings in connection with embodiments.
In order that those skilled in the art will better understand the present application, a technical solution in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in which it is apparent that the described embodiments are only some embodiments of the present application, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the present application without making any inventive effort, shall fall within the scope of the present application.
It should be noted that the terms "first," "second," and the like in the description and the claims of the present application and the above figures are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate in order to describe the embodiments of the application herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.
It should be noted that, related information (including, but not limited to, user equipment information, user personal information, etc.) and data (including, but not limited to, data for presentation, analyzed data, etc.) related to the present disclosure are information and data authorized by a user or sufficiently authorized by each party. For example, an interface is provided between the system and the relevant user or institution, before acquiring the relevant information, the system needs to send an acquisition request to the user or institution through the interface, and acquire the relevant information after receiving the consent information fed back by the user or institution.
The present application will be described with reference to preferred implementation steps, and fig. 1 is a flowchart of a recommendation method for a financial product according to an embodiment of the present application, as shown in fig. 1, and the method includes the following steps:
step S101, historical behavior data of a first object on financial products in a target database is obtained, and a target map is constructed according to the historical behavior data and all the financial products in the target database.
Optionally, the historical behavior data of the user (i.e. the first object) may be obtained from an application program corresponding to the financial product, where the historical behavior data is interaction behavior data between the user and the financial product, such as interaction behavior data including purchase, sharing, attention, and the like. And constructing a target map through the historical behavior data and all financial products in the target database. It should be noted that the nodes in the target graph represent the user and the financial product.
Step S102, processing the target map through a graph neural network model to obtain a feature vector set, wherein the feature vector set at least comprises: the first feature vector corresponding to the first object and the second feature vector corresponding to each financial product in the target database.
Optionally, the target atlas is input into a graph neural network model, and the characteristics of the target atlas are mined through the graph neural network model to obtain a first characteristic vector corresponding to the first object and a second characteristic vector corresponding to each financial product in the target database.
Step S103, the first feature vector and the second feature vector are spliced through the graph neural network model to obtain a target feature vector, and the preference score of the first object on each financial product in the target database is obtained according to the target feature vector.
Optionally, the first feature vector of the first object and the second feature vector of the product are spliced through the graphic neural network model to obtain target feature information, and then the target feature information is processed through a multi-layer perceptron layer in the graphic neural network model to obtain the preference score of the final user on the product, wherein the higher the score is, the more interested the user is in the financial product.
Step S104, screening from the financial products in the target database according to the preference scores, and determining target financial products to be recommended.
Optionally, after the preference score is obtained, the financial products in the target database are screened according to the preference score, for example, the financial products are ranked according to the preference score, and then the financial products in the first three ranks are determined as target financial products. In an alternative embodiment, the target financial product may be pushed to the target object by a corresponding application.
In summary, historical behavior data of the first object on the financial products in the target database is collected, a target map is established according to the historical behavior data and all the financial products in the target database, the association relationship between the first object and the products is described through the target map, then feature extraction is performed on the target map through a graph neural network model, a first feature vector corresponding to the first object and a second feature vector corresponding to the financial products are obtained, further preference scores of the first object on each financial product in the target database are obtained through the first feature vector and the second feature vector, finally the financial products in the target database are screened according to the preference scores, the target financial products to be recommended are determined, the association degree between users and the products can be captured more accurately through the target map, and the effect of improving the accuracy of recommending the financial products is achieved.
Optionally, in the recommendation method of financial products provided by the embodiment of the present application, constructing the target map according to the historical behavior data and all the financial products in the target database includes: determining a plurality of second objects corresponding to each financial product in the target database, wherein the second objects are objects having interaction behaviors with the financial products in the target database; acquiring first scoring data of the second object on the financial products in the target database according to the interaction behavior data between the second object and the financial products in the target database; acquiring second scoring data of the first object on financial products corresponding to the historical behavior data according to the historical behavior data; and constructing a target map according to the historical behavior data, the first scoring data and the second scoring data.
In an alternative embodiment, a plurality of second objects corresponding to each financial product in the target database are determined, where it is noted that the second objects are objects having interaction behaviors with the financial products in the target database, and the second objects include the first objects. And then determining first scoring data of the second object on the financial products interacted with the second object in the target database through interaction behavior data between the second object and the financial products in the target database. For example, different ratings are given according to different interactions, e.g., purchasing corresponding ratings > sharing corresponding ratings > focusing on corresponding ratings > clicking on corresponding ratings, etc.
And then, determining second scoring data of the first object on the financial product corresponding to the historical behavior data according to the historical behavior data of the first object on the financial product, and finally constructing a target map according to the historical behavior data, the first scoring data and the second scoring data.
The target map can accurately and comprehensively represent the user, the product and the association relation between the user and the product, and is beneficial to more accurately acquiring the feature vectors corresponding to the user and the product.
Optionally, in the recommendation method of a financial product provided by the embodiment of the present application, processing, by using a graph neural network model, a target graph to obtain a feature vector set includes: extracting features of a target map through a graph neural network model to obtain a first initial feature vector corresponding to a first object, a second initial feature vector corresponding to each financial product in a target database, a third initial feature vector and a fourth initial feature vector, wherein the third initial feature vector is a feature vector corresponding to first scoring data, and the fourth initial feature vector is a feature vector corresponding to second scoring data; determining a first feature vector according to the second initial feature vector and the third initial feature vector; determining a second feature vector according to the first initial feature vector and the fourth initial feature vector; and determining a feature vector set according to the first feature vector and the second feature vector.
In an alternative embodiment, first, the neural network model performs feature extraction on the target atlas to obtain a first initial feature vector corresponding to the first object, a second initial feature vector corresponding to each financial product in the target database, a third initial feature vector corresponding to the first scoring data, and a fourth initial feature vector corresponding to the second scoring data.
Then, the first feature vector of the first object is obtained through the second initial feature vector corresponding to each financial product and the third initial feature vector corresponding to the first scoring data, and the preference condition of the user on the financial products can be accurately represented through the third initial feature vector. And determining a second feature vector of the financial product by the first initial feature vector and the fourth initial feature vector corresponding to the first object.
And finally, obtaining the feature vector set through the first feature vector and the second feature vector.
In summary, the related characteristic information can be accurately extracted through the graph neural network model, which is beneficial to improving the accuracy of follow-up prediction on preference of financial products.
Optionally, in the recommendation method for a financial product provided by the embodiment of the present application, determining the first feature vector according to the second initial feature vector and the third initial feature vector includes: calculating according to the second initial feature vector and the third initial feature vector to obtain a first interaction feature vector of the first object on the financial product corresponding to the historical behavior data; calculating according to the interaction feature vector and the first initial feature vector to obtain target scoring data of the financial product corresponding to the historical behavior data by the first object; and determining a first feature vector according to the target scoring data.
In an alternative embodiment, determining the first feature vector from the second initial feature vector and the third initial feature vector comprises the steps of: firstly, calculating through the second initial feature vector and the third initial feature vector to obtain a first interaction feature vector of the first object on the financial product corresponding to the historical behavior data.
In an alternative embodiment, the first interaction feature vector may be obtained in the following manner:
Wherein x is a first interaction feature vector, q 1 is the second initial feature vector, e 1 is the third initial feature vector, and g v is a standard multi-layer perceptron in the graph neural network model.
And then, calculating the interaction feature vector and the first initial feature vector to obtain target scoring data of the first object on the financial product corresponding to the historical behavior data. In an alternative embodiment, the target scoring data may be obtained by:
Wherein c is target scoring data, and p is a first initial feature vector. The similarity degree between the first initial feature vector and the first interaction feature vector of the user can be calculated through the formula (2).
Finally, the first feature vector is determined according to the target scoring data, for example, the weight of the second initial feature vector can be set according to the target scoring data, and then the first feature vector is obtained by aggregation of the weight, the second initial feature vector and the first initial feature vector.
The first feature vector representing the preference of the user can be accurately obtained by aggregating the feature vectors through the steps.
Optionally, in the recommendation method for a financial product provided by the embodiment of the present application, determining the first feature vector according to the target scoring data includes: calculating according to the target scoring data to obtain a target specific gravity value of the financial product corresponding to the historical behavior data; determining a plurality of initial financial products from the financial products corresponding to the historical behavior data according to the target specific gravity value; and carrying out aggregation processing on the first interaction feature vectors corresponding to the plurality of initial financial products to obtain first feature vectors.
In an alternative embodiment, determining the first feature vector based on the target scoring data comprises the steps of: calculating the target proportion value of each target scoring data in all target scoring data according to the target scoring data, determining a plurality of initial financial products from financial products corresponding to the historical behavior data according to the target proportion value, and finally carrying out aggregation processing on first interaction feature vectors corresponding to the plurality of initial financial products to obtain first feature vectors.
In an alternative embodiment, the first feature vector may be obtained by:
h＝σ(W·{∑ixi}+b)(3)
Wherein h is a first eigenvector, sigma is a nonlinear activation function, x i is a first interaction eigenvector corresponding to the selected initial financial product, and W and b are weights and deviations of the graph neural network model.
And selecting a first interaction feature vector corresponding to a certain proportion of initial financial products through the target specific gravity value to perform subsequent neighborhood aggregation operation, so that the accuracy of the first feature vector of the user can be effectively improved.
Optionally, in the recommendation method of financial products provided by the embodiment of the present application, performing aggregation processing on first interaction feature vectors corresponding to a plurality of initial financial products, to obtain first feature vectors includes: acquiring a target weight value corresponding to each initial financial product through a multi-head attention mechanism, wherein the target weight value is used for representing the contribution degree of the first feature vector; and carrying out aggregation processing according to the target weight value corresponding to each initial financial product and the first interaction feature vector to obtain a first feature vector.
In an optional embodiment, when the first interaction feature vector is aggregated, a target weight value corresponding to each initial financial product may be obtained through a multi-head attention mechanism, where it is to be noted that the target weight value is used to characterize a contribution degree to the first feature vector, and then the first feature vector is obtained through aggregation processing of the target weight value and the first interaction feature vector.
In an alternative embodiment, the multi-head attention mechanism derives k target weight values from the bias of the weight multi-head attention mechanisms of the k sets of multi-head attention mechanisms.
In an alternative embodiment, each of the weight values calculated above requires normalization by a softmax function.
In an alternative embodiment, the first feature vector may be obtained by:
Wherein h is the first feature vector, W (k) is the weight of the multi-head attention mechanism, b is the bias, And the target weight value is normalized.
The first feature vectors obtained by setting different weight values can more accurately represent the preference of the user to the financial products.
Optionally, in the recommendation method for a financial product provided by the embodiment of the present application, determining the second feature vector according to the first initial feature vector and the fourth initial feature vector includes: calculating according to the first initial feature vector and the fourth initial feature vector to obtain a second interaction feature vector of the second object on each financial product in the target database; and determining a second feature vector according to the second interaction feature vector.
In an alternative embodiment, determining the second feature vector from the first initial feature vector and the fourth initial feature vector comprises the steps of: and calculating through the first initial feature vector and the fourth initial feature vector to obtain a second interaction feature vector of the second object on each financial product in the target database. In an alternative embodiment, the second interaction feature vector described above may be obtained in the following manner:
Where f is the second interaction feature vector, p t is the first initial feature vector, and e 2 is the fourth initial feature vector.
After the second interaction feature vector is obtained, the second interaction feature vector is used for determining the second feature vector. In an alternative embodiment, the second feature vector may be obtained in the following way:
z＝σ(W·AGG({f})+b) (6)
where z is the second eigenvector and AGG is the aggregation function.
In an alternative embodiment, RMSE (root mean square error) and MAE (mean absolute error) may be used to verify the recommendation, i.e. the recommendation ability of the graph neural network model is evaluated by the root mean square error between the actual score of the product and the predicted preference score by the user, as well as the mean absolute error.
In an alternative embodiment, accurate recommendation of financial products may be achieved through a flow chart as shown in FIG. 2: the method comprises the steps of obtaining historical behavior data of a user in an application program, giving different scores (purchase > share > attention > click) according to the historical behavior data, converting interaction between the user and a product into a user product interaction diagram, wherein nodes in the diagram represent the user and the product, and the user product interaction diagram comprises relations between the user and the product, scoring data and the like. And obtaining the feature vector corresponding to the financial product, the feature vector corresponding to the user and the grading feature vector through the user product interaction diagram. And then, respectively aggregating different feature vectors from the angles of users and products by using a multi-layer perceptron and a multi-head attention mechanism to obtain a first feature vector corresponding to the users and a second feature vector corresponding to the financial products, and finally splicing the user and the product vectors, and processing the user and the product vectors through the multi-layer perceptron layer to obtain the preference degree of the final user on the products.
According to the recommendation method of the financial products, the historical behavior data of the first object on the financial products in the target database is obtained, and the target map is constructed according to the historical behavior data and all the financial products in the target database; processing the target map through the graph neural network model to obtain a feature vector set, wherein the feature vector set at least comprises: a first feature vector corresponding to the first object and a second feature vector corresponding to each financial product in the target database; the first feature vector and the second feature vector are spliced through the graph neural network model to obtain a target feature vector, and the preference score of the first object to each financial product in the target database is obtained according to the target feature vector; the target financial products to be recommended are determined by screening from the financial products in the target database according to the preference scores, and the problem that the accuracy of the recommended financial products is low due to the fact that the financial products purchased by the user are recommended to the user in the related technology is solved. According to the technical scheme, historical behavior data of a first object on financial products in a target database are collected, a target map is established according to the historical behavior data and all the financial products in the target database, the association relation between the first object and the products is described through the target map, then feature extraction is carried out on the target map through a graph neural network model, a first feature vector corresponding to the first object and a second feature vector corresponding to the financial products are obtained, further preference scores of the first object on each financial product in the target database are obtained through the first feature vector and the second feature vector, finally the financial products to be recommended are screened from the financial products in the target database according to the preference scores, the association degree between users and the products can be captured more accurately through the target map, and the effect of improving the accuracy of recommending the financial products is achieved.
It should be noted that the steps illustrated in the flowcharts of the figures may be performed in a computer system such as a set of computer executable instructions, and that although a logical order is illustrated in the flowcharts, in some cases the steps illustrated or described may be performed in an order other than that illustrated herein.
The embodiment of the application also provides a recommending device for the financial product, and the recommending device for the financial product can be used for executing the recommending method for the financial product. The recommendation device for the financial products provided by the embodiment of the application is described below.
Fig. 3 is a schematic diagram of a recommending apparatus for financial products according to an embodiment of the present application. As shown in fig. 3, the apparatus includes: the device comprises an acquisition unit 301, a processing unit 302, a splicing unit 303 and a screening unit 304.
An obtaining unit 301, configured to obtain historical behavior data of the first object on financial products in the target database, and construct a target map according to the historical behavior data and all financial products in the target database;
The processing unit 302 is configured to process the target atlas through the graph neural network model to obtain a feature vector set, where the feature vector set at least includes: a first feature vector corresponding to the first object and a second feature vector corresponding to each financial product in the target database;
the stitching unit 303 is configured to stitch the first feature vector and the second feature vector through the neural network model to obtain a target feature vector, and obtain a preference score of the first object for each financial product in the target database according to the target feature vector;
And the screening unit 304 is configured to screen from the financial products in the target database according to the preference scores, and determine a target financial product to be recommended.
According to the recommendation device for the financial products, provided by the embodiment of the application, the historical behavior data of the first object on the financial products in the target database is obtained through the obtaining unit 301, and the target map is constructed according to the historical behavior data and all the financial products in the target database; the processing unit 302 processes the target atlas through the graph neural network model to obtain a feature vector set, where the feature vector set at least includes: a first feature vector corresponding to the first object and a second feature vector corresponding to each financial product in the target database; the splicing unit 303 performs splicing processing on the first feature vector and the second feature vector through the graph neural network model to obtain a target feature vector, and obtains a preference score of the first object on each financial product in the target database according to the target feature vector; the screening unit 304 screens from the financial products in the target database according to the preference scores to determine the target financial products to be recommended, thereby solving the problem that the accuracy of the recommended financial products is lower because the financial products purchased by the user are recommended to the user in the related art. According to the technical scheme, historical behavior data of a first object on financial products in a target database are collected, a target map is established according to the historical behavior data and all the financial products in the target database, the association relation between the first object and the products is described through the target map, then feature extraction is carried out on the target map through a graph neural network model, a first feature vector corresponding to the first object and a second feature vector corresponding to the financial products are obtained, further preference scores of the first object on each financial product in the target database are obtained through the first feature vector and the second feature vector, finally the financial products to be recommended are screened from the financial products in the target database according to the preference scores, the association degree between users and the products can be captured more accurately through the target map, and the effect of improving the accuracy of recommending the financial products is achieved.
Optionally, in the recommendation device for a financial product provided in the embodiment of the present application, the obtaining unit includes: the determining subunit is used for determining a plurality of second objects corresponding to each financial product in the target database, wherein the second objects are objects with interaction behaviors with the financial products in the target database; the first obtaining subunit is used for obtaining first scoring data of the second object on the financial products in the target database according to the interaction behavior data between the second object and the financial products in the target database; the second obtaining subunit is used for obtaining second scoring data of the financial products corresponding to the historical behavior data by the first object according to the historical behavior data; the construction unit is used for constructing a target map according to the historical behavior data, the first scoring data and the second scoring data.
Optionally, in the recommendation device for a financial product provided by the embodiment of the present application, the processing unit includes: the extraction subunit is used for carrying out feature extraction on the target map through the graph neural network model to obtain a first initial feature vector corresponding to the first object, a second initial feature vector corresponding to each financial product in the target database, a third initial feature vector and a fourth initial feature vector, wherein the third initial feature vector is a feature vector corresponding to the first scoring data, and the fourth initial feature vector is a feature vector corresponding to the second scoring data; a first determining subunit, configured to determine a first feature vector according to the second initial feature vector and the third initial feature vector; a second determining subunit, configured to determine a second feature vector according to the first initial feature vector and the fourth initial feature vector; and the third determining subunit is used for determining the feature vector set according to the first feature vector and the second feature vector.
Optionally, in the recommendation device for a financial product provided in the embodiment of the present application, the first determining subunit includes: the first computing module is used for computing according to the second initial feature vector and the third initial feature vector to obtain a first interaction feature vector of the first object on the financial product corresponding to the historical behavior data; the second calculation module is used for calculating according to the interaction feature vector and the first initial feature vector to obtain target scoring data of the financial product corresponding to the historical behavior data by the first object; and the first determining module is used for determining a first feature vector according to the target scoring data.
Optionally, in the recommendation device for a financial product provided by the embodiment of the present application, the first determining module includes: the calculation sub-module is used for calculating according to the target scoring data to obtain a target specific gravity value of the financial product corresponding to the historical behavior data; the first determining submodule is used for determining a plurality of initial financial products from financial products corresponding to the historical behavior data according to the target specific gravity value; and the aggregation sub-module is used for carrying out aggregation processing on the first interaction feature vectors corresponding to the plurality of initial financial products to obtain first feature vectors.
Optionally, in the recommendation device for a financial product provided by the embodiment of the present application, the aggregation sub-module includes: the acquisition sub-module is used for acquiring a target weight value corresponding to each initial financial product through a multi-head attention mechanism, wherein the target weight value is used for representing the contribution degree of the first feature vector; and the aggregation sub-module is used for carrying out aggregation processing according to the target weight value corresponding to each initial financial product and the first interaction feature vector to obtain a first feature vector.
Optionally, in the recommendation device for a financial product provided in the embodiment of the present application, the second determining subunit includes: the third calculation module is used for calculating according to the first initial feature vector and the fourth initial feature vector to obtain a second interaction feature vector of the second object on each financial product in the target database; and the second determining module is used for determining a second characteristic vector according to the second interaction characteristic vector.
The recommendation device for financial products comprises a processor and a memory, wherein the acquisition unit 301, the processing unit 302, the splicing unit 303, the screening unit 304 and the like are stored in the memory as program units, and the processor executes the program units stored in the memory to realize corresponding functions.
The processor includes a kernel, and the kernel fetches the corresponding program unit from the memory. The kernel can be provided with one or more, and accurate recommendation of financial products is realized by adjusting kernel parameters.
The memory may include volatile memory, random Access Memory (RAM), and/or nonvolatile memory, such as Read Only Memory (ROM) or flash memory (flash RAM), among other forms in computer readable media, the memory including at least one memory chip.
An embodiment of the present invention provides a computer-readable storage medium having a program stored thereon, which when executed by a processor, implements a recommendation method for a financial product.
The embodiment of the invention provides a processor, which is used for running a program, wherein the program runs to execute a recommendation method of financial products.
As shown in fig. 4, an embodiment of the present invention provides an electronic device, where the device includes a processor, a memory, and a program stored in the memory and executable on the processor, and when the processor executes the program, the following steps are implemented: acquiring historical behavior data of a first object on financial products in a target database, and constructing a target map according to the historical behavior data and all the financial products in the target database; processing the target map through the graph neural network model to obtain a feature vector set, wherein the feature vector set at least comprises: a first feature vector corresponding to the first object and a second feature vector corresponding to each financial product in the target database; the first feature vector and the second feature vector are spliced through the graph neural network model to obtain a target feature vector, and the preference score of the first object to each financial product in the target database is obtained according to the target feature vector; and screening from the financial products in the target database according to the preference scores to determine target financial products to be recommended.
Optionally, constructing the target atlas according to the historical behavior data and all financial products in the target database includes: determining a plurality of second objects corresponding to each financial product in the target database, wherein the second objects are objects having interaction behaviors with the financial products in the target database; acquiring first scoring data of the second object on the financial products in the target database according to the interaction behavior data between the second object and the financial products in the target database; acquiring second scoring data of the first object on financial products corresponding to the historical behavior data according to the historical behavior data; and constructing a target map according to the historical behavior data, the first scoring data and the second scoring data.
Optionally, processing the target atlas through the graph neural network model, and obtaining the feature vector set includes: extracting features of a target map through a graph neural network model to obtain a first initial feature vector corresponding to a first object, a second initial feature vector corresponding to each financial product in a target database, a third initial feature vector and a fourth initial feature vector, wherein the third initial feature vector is a feature vector corresponding to first scoring data, and the fourth initial feature vector is a feature vector corresponding to second scoring data; determining a first feature vector according to the second initial feature vector and the third initial feature vector; determining a second feature vector according to the first initial feature vector and the fourth initial feature vector; and determining a feature vector set according to the first feature vector and the second feature vector.
Optionally, determining the first feature vector according to the second initial feature vector and the third initial feature vector includes: calculating according to the second initial feature vector and the third initial feature vector to obtain a first interaction feature vector of the first object on the financial product corresponding to the historical behavior data; calculating according to the interaction feature vector and the first initial feature vector to obtain target scoring data of the financial product corresponding to the historical behavior data by the first object; and determining a first feature vector according to the target scoring data.
Optionally, determining the first feature vector according to the target scoring data includes: calculating according to the target scoring data to obtain a target specific gravity value of the financial product corresponding to the historical behavior data; determining a plurality of initial financial products from the financial products corresponding to the historical behavior data according to the target specific gravity value; and carrying out aggregation processing on the first interaction feature vectors corresponding to the plurality of initial financial products to obtain first feature vectors.
Optionally, performing aggregation processing on first interaction feature vectors corresponding to the plurality of initial financial products, to obtain first feature vectors, including: acquiring a target weight value corresponding to each initial financial product through a multi-head attention mechanism, wherein the target weight value is used for representing the contribution degree of the first feature vector; and carrying out aggregation processing according to the target weight value corresponding to each initial financial product and the first interaction feature vector to obtain a first feature vector.
Optionally, determining the second feature vector according to the first initial feature vector and the fourth initial feature vector includes: calculating according to the first initial feature vector and the fourth initial feature vector to obtain a second interaction feature vector of the second object on each financial product in the target database; and determining a second feature vector according to the second interaction feature vector.
The device herein may be a server, PC, PAD, cell phone, etc.
The application also provides a computer program product adapted to perform, when executed on a data processing device, a program initialized with the method steps of: acquiring historical behavior data of a first object on financial products in a target database, and constructing a target map according to the historical behavior data and all the financial products in the target database; processing the target map through the graph neural network model to obtain a feature vector set, wherein the feature vector set at least comprises: a first feature vector corresponding to the first object and a second feature vector corresponding to each financial product in the target database; the first feature vector and the second feature vector are spliced through the graph neural network model to obtain a target feature vector, and the preference score of the first object to each financial product in the target database is obtained according to the target feature vector; and screening from the financial products in the target database according to the preference scores to determine target financial products to be recommended.
Optionally, constructing the target atlas according to the historical behavior data and all financial products in the target database includes: determining a plurality of second objects corresponding to each financial product in the target database, wherein the second objects are objects having interaction behaviors with the financial products in the target database; acquiring first scoring data of the second object on the financial products in the target database according to the interaction behavior data between the second object and the financial products in the target database; acquiring second scoring data of the first object on financial products corresponding to the historical behavior data according to the historical behavior data; and constructing a target map according to the historical behavior data, the first scoring data and the second scoring data.
Optionally, processing the target atlas through the graph neural network model, and obtaining the feature vector set includes: extracting features of a target map through a graph neural network model to obtain a first initial feature vector corresponding to a first object, a second initial feature vector corresponding to each financial product in a target database, a third initial feature vector and a fourth initial feature vector, wherein the third initial feature vector is a feature vector corresponding to first scoring data, and the fourth initial feature vector is a feature vector corresponding to second scoring data; determining a first feature vector according to the second initial feature vector and the third initial feature vector; determining a second feature vector according to the first initial feature vector and the fourth initial feature vector; and determining a feature vector set according to the first feature vector and the second feature vector.
Optionally, determining the first feature vector according to the second initial feature vector and the third initial feature vector includes: calculating according to the second initial feature vector and the third initial feature vector to obtain a first interaction feature vector of the first object on the financial product corresponding to the historical behavior data; calculating according to the interaction feature vector and the first initial feature vector to obtain target scoring data of the financial product corresponding to the historical behavior data by the first object; and determining a first feature vector according to the target scoring data.
Optionally, determining the first feature vector according to the target scoring data includes: calculating according to the target scoring data to obtain a target specific gravity value of the financial product corresponding to the historical behavior data; determining a plurality of initial financial products from the financial products corresponding to the historical behavior data according to the target specific gravity value; and carrying out aggregation processing on the first interaction feature vectors corresponding to the plurality of initial financial products to obtain first feature vectors.
Optionally, performing aggregation processing on first interaction feature vectors corresponding to the plurality of initial financial products, to obtain first feature vectors, including: acquiring a target weight value corresponding to each initial financial product through a multi-head attention mechanism, wherein the target weight value is used for representing the contribution degree of the first feature vector; and carrying out aggregation processing according to the target weight value corresponding to each initial financial product and the first interaction feature vector to obtain a first feature vector.
Optionally, determining the second feature vector according to the first initial feature vector and the fourth initial feature vector includes: calculating according to the first initial feature vector and the fourth initial feature vector to obtain a second interaction feature vector of the second object on each financial product in the target database; and determining a second feature vector according to the second interaction feature vector.
It will be appreciated by those skilled in the art that embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
In one typical configuration, a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.
The memory may include volatile memory in a computer-readable medium, random Access Memory (RAM) and/or nonvolatile memory, etc., such as Read Only Memory (ROM) or flash RAM. Memory is an example of a computer-readable medium.
Computer readable media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of storage media for a computer include, but are not limited to, phase change memory (PRAM), static Random Access Memory (SRAM), dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), read Only Memory (ROM), electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape disk storage or other magnetic storage devices, or any other non-transmission medium, which can be used to store information that can be accessed by a computing device. Computer-readable media, as defined herein, does not include transitory computer-readable media (transmission media), such as modulated data signals and carrier waves.
It should also be noted that the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises an element.
It will be appreciated by those skilled in the art that embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The foregoing is merely exemplary of the present application and is not intended to limit the present application. Various modifications and variations of the present application will be apparent to those skilled in the art. Any modification, equivalent replacement, improvement, etc. which come within the spirit and principles of the application are to be included in the scope of the claims of the present application.

Claims (10)

1. A recommendation method for a financial product, comprising:
acquiring historical behavior data of a first object on financial products in a target database, and constructing a target map according to the historical behavior data and all the financial products in the target database;
processing the target map through a graph neural network model to obtain a feature vector set, wherein the feature vector set at least comprises: a first feature vector corresponding to the first object and a second feature vector corresponding to each financial product in the target database;
The first feature vector and the second feature vector are spliced through the graph neural network model to obtain a target feature vector, and the preference score of the first object to each financial product in a target database is obtained according to the target feature vector;
And screening from the financial products in the target database according to the preference scores to determine target financial products to be recommended.
2. The method of claim 1, wherein constructing a target profile from the historical behavioral data and all financial products in a target database comprises:
Determining a plurality of second objects corresponding to each financial product in the target database, wherein the second objects are objects having interaction behaviors with the financial products in the target database;
acquiring first scoring data of the second object on the financial products in the target database according to the interaction behavior data between the second object and the financial products in the target database;
acquiring second scoring data of the first object on financial products corresponding to the historical behavior data according to the historical behavior data;
And constructing the target map according to the historical behavior data, the first scoring data and the second scoring data.
3. The method of claim 2, wherein processing the target atlas through a graph neural network model to obtain a set of feature vectors comprises:
Extracting features of the target map through the graph neural network model to obtain a first initial feature vector corresponding to the first object, a second initial feature vector corresponding to each financial product in the target database, a third initial feature vector and a fourth initial feature vector, wherein the third initial feature vector is a feature vector corresponding to the first scoring data, and the fourth initial feature vector is a feature vector corresponding to the second scoring data;
Determining the first feature vector according to the second initial feature vector and the third initial feature vector;
determining the second feature vector according to the first initial feature vector and the fourth initial feature vector;
and determining the feature vector set according to the first feature vector and the second feature vector.
4. A method according to claim 3, wherein determining the first feature vector from the second initial feature vector and the third initial feature vector comprises:
Calculating according to the second initial feature vector and the third initial feature vector to obtain a first interaction feature vector of the first object on the financial product corresponding to the historical behavior data;
Calculating according to the interaction feature vector and the first initial feature vector to obtain target scoring data of the first object on the financial product corresponding to the historical behavior data;
and determining the first feature vector according to the target scoring data.
5. The method of claim 4, wherein determining the first feature vector based on the target scoring data comprises:
Calculating according to the target scoring data to obtain a target proportion value of the financial product corresponding to the historical behavior data;
Determining a plurality of initial financial products from the financial products corresponding to the historical behavior data according to the target specific gravity value;
And carrying out aggregation processing on the first interaction feature vectors corresponding to the plurality of initial financial products to obtain the first feature vectors.
6. The method of claim 5, wherein aggregating the first interaction feature vectors corresponding to the plurality of initial financial products to obtain the first feature vectors comprises:
Acquiring a target weight value corresponding to each initial financial product through a multi-head attention mechanism, wherein the target weight value is used for representing the contribution degree of the first feature vector;
and carrying out aggregation processing according to the target weight value corresponding to each initial financial product and the first interaction feature vector to obtain the first feature vector.
7. A method according to claim 3, wherein determining the second feature vector from the first initial feature vector and the fourth initial feature vector comprises:
Calculating according to the first initial feature vector and the fourth initial feature vector to obtain a second interaction feature vector of the second object on each financial product in the target database;
and determining the second feature vector according to the second interaction feature vector.
8. A recommendation device for a financial product, comprising:
The acquisition unit is used for acquiring historical behavior data of the first object on the financial products in the target database and constructing a target map according to the historical behavior data and all the financial products in the target database;
The processing unit is used for processing the target map through a graph neural network model to obtain a feature vector set, wherein the feature vector set at least comprises: a first feature vector corresponding to the first object and a second feature vector corresponding to each financial product in the target database;
The splicing unit is used for carrying out splicing processing on the first feature vector and the second feature vector through the graph neural network model to obtain a target feature vector, and obtaining the preference score of the first object on each financial product in a target database according to the target feature vector;
And the screening unit is used for screening from the financial products in the target database according to the preference scores and determining target financial products to be recommended.
9. A computer-readable storage medium, characterized in that the computer-readable storage medium comprises a stored program, wherein the program, when run, controls the storage medium to perform the recommendation method of the financial product of any one of claims 1 to 7 at a device.
10. An electronic device comprising one or more processors and a memory for storing one or more programs, wherein the one or more programs, when executed by the one or more processors, cause the one or more processors to implement the method of recommending financial products of any of claims 1-7.
CN202410177654.5A 2024-02-08 2024-02-08 Recommendation method and device for financial products, storage medium and electronic equipment Pending CN117952719A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202410177654.5A CN117952719A (en) 2024-02-08 2024-02-08 Recommendation method and device for financial products, storage medium and electronic equipment

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202410177654.5A CN117952719A (en) 2024-02-08 2024-02-08 Recommendation method and device for financial products, storage medium and electronic equipment

Publications (1)

Publication Number Publication Date
CN117952719A true CN117952719A (en) 2024-04-30

Family

ID=90801425

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202410177654.5A Pending CN117952719A (en) 2024-02-08 2024-02-08 Recommendation method and device for financial products, storage medium and electronic equipment

Country Status (1)

Country Link
CN (1) CN117952719A (en)

Similar Documents

Publication Publication Date Title
US11122333B2 (en) User feature generation method and apparatus, device, and computer-readable storage medium
CN108829808B (en) Page personalized sorting method and device and electronic equipment
US20200210708A1 (en) Method and device for video classification
US20180240041A1 (en) Distributed hyperparameter tuning system for machine learning
CN110008397B (en) Recommendation model training method and device
CN114359563B (en) Model training method, device, computer equipment and storage medium
CN108540860B (en) Video recall method and device
Luo et al. Online optimization of collaborative web service QoS prediction based on approximate dynamic programming
CN111079944A (en) Method and device for realizing interpretation of transfer learning model, electronic equipment and storage medium
CN116932909A (en) Model recommendation method and device, processor and electronic equipment
CN111159563A (en) Method, device and equipment for determining user interest point information and storage medium
CN117520645A (en) User determination method and device based on financial products and electronic equipment
Cuevas et al. Otsu and Kapur segmentation based on harmony search optimization
CN116205737A (en) Data analysis method applied to financial system and cloud server
CN117952719A (en) Recommendation method and device for financial products, storage medium and electronic equipment
CN114897607A (en) Data processing method and device for product resources, electronic equipment and storage medium
CN115080856A (en) Recommendation method and device and training method and device of recommendation model
CN117726403A (en) Product recommendation method and device, storage medium and electronic equipment
CN112348583B (en) User preference generation method and generation system
US20240201957A1 (en) Neural network model definition code generation and optimization
CN114580427B (en) Self-media user selection method and related equipment
CN116561735B (en) Mutual trust authentication method and system based on multiple authentication sources and electronic equipment
US20240256981A1 (en) Self-adaptive multi-model approach in representation feature space for propensity to action
CN116739708A (en) Product recommendation method and device, processor and electronic equipment
CN118446744A (en) Rights and interests data conversion method and system based on artificial intelligence

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination