Product recommendation method and apparatus, storage medium, and electronic device

Abstract

This application discloses a product recommendation method, apparatus, storage medium, and electronic device. Relating to fintech or other related fields, the method includes: acquiring N first financial products associated with a target user; for each first financial product, determining the influence of the first financial product based on a target heterogeneous information network, wherein the target heterogeneous information network includes at least the following types of nodes: user type, product type, and edges between nodes to represent the association relationships between nodes; calculating the similarity between the first financial product and each second financial product based on the target heterogeneous information network; and based on the influence of the N first financial products and the similarity between the N first financial products and each second financial product, determining a target financial product from the second financial products and recommending the target financial product to the target user. This application solves the problem of low accuracy in product recommendations in related technologies.

Description

Technical Field

[0001] This application relates to financial technology or other related fields, and more specifically, to a product recommendation method, apparatus, storage medium, and electronic device. Background Technology

[0002] With the continuous development of technology and the digital economy, financial product recommendations have become an important means for institutions across various sectors to enhance user experience and market competitiveness. Whether in e-commerce, online media, or the financial services industry, efficient and accurate product recommendations can effectively enhance user satisfaction and stickiness, while creating more business opportunities and revenue for institutions.

[0003] Currently, collaborative filtering and content-based recommendation methods are commonly used in related technologies. These methods typically only consider the product's own attributes (such as product type) or the user's past behavior, resulting in low recommendation accuracy.

[0004] There is currently no effective solution to the aforementioned problems in the relevant technologies. Summary of the Invention

[0005] The main objective of this application is to provide a product recommendation method, apparatus, storage medium, and electronic device to solve the problem of low accuracy in product recommendations in related technologies.

[0006] To achieve the above objectives, according to one aspect of this application, a product recommendation method is provided. The method includes: obtaining N first financial products associated with a target user, where N is a positive integer greater than 1; for each first financial product, determining the influence of the first financial product based on a target heterogeneous information network, where influence characterizes the importance of the first financial product, the target heterogeneous information network includes at least the following types of nodes: user type, product type, and edges between nodes are used to characterize the association relationship between nodes, the target heterogeneous information network includes first financial products and second financial products; calculating the similarity between the first financial product and each of the second financial products based on the target heterogeneous information network; and determining a target financial product from the second financial products based on the influence of the N first financial products and the similarity between the N first financial products and each of the second financial products, and recommending the target financial product to the target user.

[0007] Optionally, the product recommendation method further includes: determining at least one network path conforming to the target meta-path from the target heterogeneous information network, wherein the target meta-path is a path template represented by node type, the node types of the start and end points of the target meta-path are product types, and the network path is formed by connecting nodes in the target heterogeneous information network; determining the influence of the first financial product based on the number of products corresponding to the starting node in the at least one network path, the influence of the target neighbor nodes, and the number of nodes associated with the target neighbor nodes, wherein the target neighbor nodes are the neighbor nodes of the first financial product in the at least one network path.

[0008] Optionally, the product recommendation method may further include: obtaining multiple candidate meta-paths before determining the influence of the first financial product based on the target heterogeneous information network; determining the sum of the influence of product type nodes in the target heterogeneous information network based on the candidate meta-paths; and determining the target meta-path from the multiple candidate meta-paths based on the sum of the influence corresponding to the multiple candidate meta-paths.

[0009] Optionally, the product recommendation method further includes: for each second financial product, determining a first network path from the target heterogeneous information network that starts with the first financial product and ends with the second financial product, and conforms to the target meta-path; determining a second network path from the target heterogeneous information network that starts with the first financial product and ends with the second financial product, and conforms to the target meta-path; determining a third network path from the target heterogeneous information network that starts with the second financial product and ends with the second financial product, and conforms to the target meta-path; and determining the similarity between the first financial product and the second financial product based on the number of paths in the first network path, the number of paths in the second network path, and the number of paths in the third network path.

[0010] Optionally, the product recommendation method further includes: for each second financial product, determining the target similarity between the second financial product and the first financial product based on the similarity between the second financial product and the first financial product and the influence of the first financial product; summing the target similarities between the second financial product and each first financial product to obtain the recommendation degree value of the second financial product; and determining the target financial product from the second financial products based on the recommendation degree value of the second financial product.

[0011] Optionally, the product recommendation method may also include: for each first financial product, normalizing the influence of the first financial product based on the sum of the influences of N first financial products to obtain the target influence of the first financial product; calculating the product of the similarity between the second financial product and the first financial product and the target influence of the first financial product to obtain the target similarity.

[0012] Optionally, the product recommendation method may also include: identifying financial products already purchased by the target user as the first financial product, and / or identifying the first financial product from among the financial products based on the interaction information between the target user and the financial products.

[0013] To achieve the above objectives, according to another aspect of this application, a product recommendation device is provided. The device includes: a first acquisition module, configured to acquire N first financial products associated with a target user, where N is a positive integer greater than 1; a first determination module, configured to determine the influence of each first financial product based on a target heterogeneous information network, wherein influence characterizes the importance of the first financial product, the target heterogeneous information network includes at least the following types of nodes: user type, product type, and edges between nodes characterize the association relationship between nodes, the target heterogeneous information network includes first financial products and second financial products; a first calculation module, configured to calculate the similarity between the first financial product and each of the second financial products based on the target heterogeneous information network; and a processing module, configured to determine a target financial product from the second financial products based on the influence of the N first financial products and the similarity between the N first financial products and each of the second financial products, and recommend the target financial product to the target user.

[0014] Optionally, the first determining module further includes: a first determining submodule, used to determine at least one network path conforming to the target meta-path from the target heterogeneous information network, wherein the target meta-path is a path template represented by node type, the node types of the starting point and the ending point of the target meta-path are product types, and the network path is formed by connecting nodes in the target heterogeneous information network; and a second determining submodule, used to determine the influence of the first financial product based on the number of products corresponding to the starting node in the at least one network path, the influence of the target neighbor node, and the number of nodes associated with the target neighbor node, wherein the target neighbor node is the neighbor node of the first financial product in the at least one network path.

[0015] Optionally, the product recommendation device further includes: a second acquisition module for acquiring multiple candidate meta-paths; a second calculation module for determining the total influence of nodes of product type in the target heterogeneous information network based on the candidate meta-paths; and a second determination module for determining the target meta-path from the multiple candidate meta-paths based on the total influence corresponding to the multiple candidate meta-paths.

[0016] Optionally, the first calculation module further includes: a third determining submodule, used to determine, for each second financial product, a network path from the target heterogeneous information network that starts with the first financial product and ends with the second financial product, and conforms to the target meta-path, to obtain a first network path; a fourth determining submodule, used to determine, from the target heterogeneous information network, a network path that starts with the first financial product and ends with the first financial product, and conforms to the target meta-path, to obtain a second network path; a fifth determining submodule, used to determine, from the target heterogeneous information network, a network path that starts with the second financial product and ends with the second financial product, and conforms to the target meta-path, to obtain a third network path; and a sixth determining submodule, used to determine the similarity between the first financial product and the second financial product based on the number of paths in the first network path, the number of paths in the second network path, and the number of paths in the third network path.

[0017] Optionally, the processing module further includes: a seventh determining submodule, used to determine the target similarity between the second financial product and the first financial product for each second financial product based on the similarity between the second financial product and the first financial product and the influence of the first financial product; a calculation submodule, used to sum the target similarities between the second financial product and each first financial product to obtain the recommendation degree value of the second financial product; and an eighth determining submodule, used to determine the target financial product from the second financial products based on the recommendation degree value of the second financial product.

[0018] Optionally, the seventh determining submodule further includes: a processing unit, used to normalize the influence of each first financial product based on the sum of the influences of N first financial products to obtain the target influence of the first financial product; and a calculation unit, used to calculate the product between the similarity between the second financial product and the first financial product and the target influence of the first financial product to obtain the target similarity.

[0019] Optionally, the first acquisition module further includes: a ninth determining submodule, used to determine the financial products purchased by the target user as the first financial product, and / or a tenth determining submodule, used to determine the first financial product from the financial products based on the interaction information between the target user and the financial products.

[0020] To achieve the above objectives, according to another aspect of this application, a computer-readable storage medium is provided, the computer-readable storage medium including a stored executable program, wherein, when the executable program is executed, the device on which the computer-readable storage medium is located controls the execution of the above-described product recommendation method.

[0021] To achieve the above objectives, according to another aspect of this application, an electronic device is provided, the electronic device including a memory storing an executable program; and a processor for running the program, wherein the program executes the Z method described above when it runs.

[0022] To achieve the above objectives, according to another aspect of this application, a computer program product is provided, including computer instructions that, when executed by a processor, implement the steps of the product recommendation method described above.

[0023] In this embodiment, by constructing a target heterogeneous information network, the association between users and financial products is comprehensively considered. The influence of the first financial product associated with the target user is determined based on the target heterogeneous information network, and the similarity between the first financial product and other financial products (i.e., each second financial product) is determined. This realizes the comprehensive association relationship in the target heterogeneous information network, determines the importance of the first financial product that the user is concerned about and the degree of association between the first financial product and other financial products, thereby effectively improving the accuracy of product recommendation when determining the target financial product based on influence and similarity.

[0024] Therefore, the method provided in this application achieves the goal of calculating the influence and similarity of products based on the target heterogeneous information network, and making product recommendations based on influence and similarity, thereby improving the technical effect of improving the accuracy of product recommendations and solving the technical problem of low accuracy of product recommendations in related technologies. Attached Figure Description

[0025] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments and descriptions of this application are used to explain this application and do not constitute an undue limitation of this application. In the drawings:

[0026] Figure 1 This is a hardware structure block diagram of a computer terminal provided according to an embodiment of this application;

[0027] Figure 2 This is a flowchart of a product recommendation method provided according to an embodiment of this application;

[0028] Figure 3 This is a schematic diagram of a target heterogeneous information network provided according to an embodiment of this application;

[0029] Figure 4 This is a schematic diagram of the node relationships in a target heterogeneous information network according to an embodiment of this application;

[0030] Figure 5 This is a schematic diagram of the target processing system provided according to an embodiment of this application;

[0031] Figure 6This is a schematic diagram of a product recommendation device provided according to an embodiment of this application;

[0032] Figure 7 This is a structural block diagram of an electronic device according to an embodiment of this application. Detailed Implementation

[0033] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.

[0034] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0035] It should be noted that the product recommendation methods, apparatus, storage media and electronic devices disclosed herein can be used in the fintech field, or in any field other than fintech. The application fields of the product recommendation methods, apparatus, storage media and electronic devices disclosed herein are not limited.

[0036] It should be noted that the information collected in this application (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for display, data used for analysis, etc.) are information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, storage, use, processing, transmission, provision, disclosure, and application of this data all comply with relevant laws, regulations, and standards, necessary confidentiality measures have been taken, and they do not violate public order and good morals. Corresponding access points are provided for users to choose to authorize or refuse. For example, interfaces are set up between this system and relevant users or organizations, providing users with corresponding access points to choose to agree to or refuse automated decision-making results; if the user chooses to refuse, the process proceeds to the expert decision-making stage.

[0037] Example 1

[0038] According to an embodiment of this application, an embodiment of a product recommendation method is also provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0039] The method embodiment provided in Embodiment 1 of this application can be executed on a mobile terminal, computer terminal, or similar computing device. Figure 1 A hardware block diagram of a computer terminal (or mobile device) for implementing a product recommendation method is shown. Figure 1 As shown, the computer terminal 10 (or mobile device) may include one or more processors 102 (shown as 102a, 102b, ..., 102n in the figure) 102 (processor 102 may include, but is not limited to, a microprocessor (MCU) or a field-programmable gate array (FPGA), etc.), a memory 104 for storing data, and a transmission device 106 for communication functions. In addition, it may also include: a display, an input / output (I / O) interface, a Universal Serial Bus (USB) port (which may be included as one of the ports of a BUS bus), a network interface, a power supply, and / or a camera. Those skilled in the art will understand that... Figure 1 The structure shown is for illustrative purposes only and does not limit the structure of the aforementioned electronic device. For example, computer terminal 10 may also include... Figure 1 The more or fewer components shown, or having the same Figure 1 The different configurations shown.

[0040] It should be noted that the aforementioned one or more processors 102 and / or other data processing circuits are generally referred to herein as "data processing circuits". These data processing circuits may be embodied, in whole or in part, in software, hardware, firmware, or any other combination thereof. Furthermore, the data processing circuits may be a single, independent processing module, or may be integrated, in whole or in part, into any other element within the computer terminal 10 (or mobile device). As involved in the embodiments of this application, the data processing circuits serve as a processor control mechanism (e.g., selection of a variable resistor termination path connected to an interface).

[0041] The memory 104 can be used to store software programs and modules of application software, such as the program instructions / data storage device corresponding to the product recommendation method in this embodiment. The processor 102 executes various functional applications and data processing by running the software programs and modules stored in the memory 104, thereby realizing the product recommendation method described above. The memory 104 may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 104 may further include memory remotely located relative to the processor 102, and these remote memories can be connected to the computer terminal 10 via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0042] The transmission device 106 is used to receive or send data via a network. Specific examples of the network described above may include a wireless network provided by the communication provider of the computer terminal 10. In one example, the transmission device 106 includes a Network Interface Controller (NIC), which can connect to other network devices via a base station to communicate with the Internet. In another example, the transmission device 106 may be a Radio Frequency (RF) module, used for wireless communication with the Internet.

[0043] The display can be, for example, a touchscreen liquid crystal display (LCD), which allows the user to interact with the user interface of the computer terminal 10 (or mobile device).

[0044] Under the aforementioned operating environment, this application provides the following: Figure 2 The product recommendation method shown. Figure 2 This is a flowchart of the product recommendation method according to Embodiment 1 of this application.

[0045] Step S201: Obtain the N first financial products associated with the target user, where N is a positive integer greater than 1.

[0046] Optionally, electronic devices, application systems, servers, or other similar devices can be used as the executing entity of this application. In this embodiment, the target processing system is used as the executing entity to execute the above-described product recommendation method. The target processing system can be a system within a financial institution.

[0047] In an optional embodiment, the target user refers to a specific user who requires personalized product recommendations; for example, the target user may be a customer of a financial institution. The first financial product refers to a financial product that has some form of association with the target user (such as purchase, inquiry, following, etc.).

[0048] Optionally, the target processing system can acquire interaction information between the target user and financial products offered by financial institutions, and then determine N primary financial products from these products based on this interaction information. For example, by analyzing the target user's transaction records, browsing history, search queries, and other behavioral data, the system can extract the N financial products that the user has interacted with most recently or most frequently as the primary financial products. In other words, the primary financial products are those that the target user is interested in (or prefers).

[0049] Step S202: For each first financial product, determine the influence of the first financial product based on the target heterogeneous information network, where influence represents the importance of the first financial product. The target heterogeneous information network includes at least the following types of nodes: user type, product type, and the edges between nodes are used to represent the association between nodes. The target heterogeneous information network includes the first financial product and the second financial product.

[0050] A heterogeneous information network (HIN) is a network structure composed of multiple types of nodes and edges, where each type of node represents a different entity or object, and each type of edge represents a different relationship between entities. Compared with traditional homogeneous information networks (such as those with a single type of node and edge), heterogeneous information networks, by introducing diverse node and edge types, can better represent the complex multidimensional relationships and structures in the real world.

[0051] In an optional embodiment, the target heterogeneous information network includes at least the following types of nodes: user type and product type. Nodes of the user type can represent different users (i.e., customers) within a financial institution, and nodes of the product type can represent different financial products within the financial institution. For example, the target heterogeneous information network is constructed by comprehensively analyzing the financial products and user purchase history within a financial institution. The target heterogeneous information network is used to describe at least the relationships between multiple users and multiple financial products within a financial institution. The relationships between users and financial products can be of various types, such as holding, purchasing, and selling. For example, user purchase, query, and following behaviors are considered as edges in the network, and basic information about users and financial products is used as node attributes. Where an edge exists between a user and a financial product, this edge is a bidirectional edge.

[0052] In an alternative embodiment, within the target heterogeneous information network, similar users can be identified based on the first financial product associated with each user, and edges can be constructed between these similar users. For example, Figure 3 This is a schematic diagram of a target heterogeneous information network provided according to an embodiment of this application, such as... Figure 3As shown, the target heterogeneous information network includes user nodes and financial product nodes. Bidirectional edges can be established between user nodes and between user nodes and financial product nodes. For example, Figure 4 This is a schematic diagram of node relationships in a target heterogeneous information network according to an embodiment of this application, such as... Figure 4 As shown, the relationship between users and financial products can be purchase, holding, etc., and the attributes of the edges can include the duration of the user's holding of the product. The target processing system can identify similar users through a homogeneous user screening mechanism and establish edges between users based on this.

[0053] In an optional embodiment, the target heterogeneous information network may include nodes of user type and product type, as well as nodes of business personnel type. For example, nodes of business personnel type are used to represent different business personnel in a financial institution. Different business personnel are responsible for maintaining different financial products. When a user purchases a financial product, edges can be established in the network between the user and the financial product, and between the user and the business personnel who maintain the financial product.

[0054] In an optional embodiment, the target heterogeneous information network may include nodes for user type and product type, as well as nodes for product tag type. For example, nodes for product tag type are used to represent tags that financial products are marked with, and when a financial product has a certain tag, an edge is established between the financial product and the tag.

[0055] Optionally, the target processing system can construct a heterogeneous information network of targets based on data from the financial institution's database. For example, after acquiring data from the database, the data can be preprocessed as follows:

[0056] (1) Regarding primary keys, delete the technical primary key and retain the business primary key. The technical primary key is a unique identifier automatically generated by the system for internal data management, while the business primary key is an identifier defined according to the actual business logic and can reflect the true characteristics of the business entity. In the process of building a heterogeneous information network, the business primary key is retained because it can better reflect the relationship between business entities, while deleting the technical primary key can avoid unnecessary noise and redundancy, optimize the network structure, and make the network more focused on business logic.

[0057] (2) For default values, delete data with default values. Default values ​​are fixed values ​​that the system automatically fills in when no values ​​are explicitly given for certain fields in the database table. These default values ​​may not reflect real transactions or user behavior. If they are directly used in network construction, they may mislead similarity analysis and influence calculation. Therefore, these data with default values ​​can be deleted in the preprocessing stage to improve the reliability of the final generated network.

[0058] (3) For duplicate data, pull duplicate data from the denormalized table into a separate node. When constructing HIN, the presence of duplicate data will affect the efficiency and performance of the network and may also lead to bias in the analysis results. Extracting duplicate data from the denormalized table and creating a separate node can avoid the same entity being represented multiple times, reduce redundant nodes in the network, and facilitate subsequent meta-path analysis and similarity calculation, thereby improving the accuracy and response speed of recommendations.

[0059] Optionally, after data preprocessing is completed, relational mapping is performed based on the relationships between the data, that is, the database relational model is converted into a target heterogeneous information network. This process involves mapping tables, rows, columns, foreign keys, and table joins in the relational database to nodes, node labels, node attributes, edges (relationships), and relational attributes in the target heterogeneous information network. For example, the target processing system can perform the following processing:

[0060] (1) Table to Node Labels: Each entity table in the relational model becomes a label on a node in the target heterogeneous information network. In a relational database, each table represents an entity type, such as "user" or "product". In the target heterogeneous information network, these entity types are represented as node labels, that is, nodes in the network are labeled to distinguish different entity types. For example, each row in the "user" table will become a node labeled "user" in the target heterogeneous information network.

[0061] (2) Row to Node: Each row in the relation becomes a node in the target heterogeneous information network. Each row of data in the relational database represents an instance of an entity. When transformed into the target heterogeneous information network, these instances will be regarded as nodes in the network, and each node represents a specific entity, such as a specific user or financial product.

[0062] (3) Column to Node Attributes: Columns (fields) in the relational table become node attributes in the target heterogeneous information network. Columns in a relational database contain specific attributes of entities, such as a user's transaction history or product type. In the target heterogeneous information network, these attributes are mapped to attributes on nodes to describe the detailed characteristics of nodes in the network.

[0063] (4) Foreign Key to Relationship: Replacing the foreign key of another table with a network relationship (edge). In relational databases, foreign keys are used to establish relationships between two tables, that is, the reference relationship between a column of one table and the primary key of another table. In the target heterogeneous information network, this reference relationship will be converted into edges (relationships) between nodes, representing the association between entities. For example, if the foreign key in the "User" table references the primary key in the "Product" table, then in the target heterogeneous information network, this will be represented as a "purchase" or "hold" relationship between the "User" node and the "Product" node.

[0064] (5) Table Joining to Relationships: This transforms the joined tables into new relationships within the target heterogeneous information network, where the columns on these tables become relational attributes. In relational databases, two or more tables can merge related data through join operations. In the target heterogeneous information network, this join operation is viewed as a complex relationship between entities. A new relation type can be created to represent this join, where the columns (fields) on the original tables are mapped to attributes of this new relation, used to describe in detail the specific association between these entities.

[0065] In an optional embodiment, after determining the target heterogeneous information network, path instances conforming to a predefined target meta-path (e.g., financial product-user-financial product) can be found within the network. Then, the influence of the first financial product is calculated based on the relationship between the first financial product and other nodes within the path instances. The influence of the first financial product can be considered its "popularity," "attractiveness," or "centrality" within the target heterogeneous information network. By calculating influence, it is possible to identify which products hold a core position in the network and which are favored by more users (more representative of public preference), thus providing a standard for the recommendation system to measure product importance. When recommending financial products, products with greater influence tend to have higher priority, which helps the system filter out products that are more popular in the market or more likely to attract specific users, thereby improving the accuracy and personalization of recommendations.

[0066] Step S203: Calculate the similarity between the first financial product and each of the second financial products based on the target heterogeneous information network.

[0067] Optionally, the second financial product can be understood as a financial product other than the first financial product in the target heterogeneous information network.

[0068] For example, the target processing system can use predefined target meta-paths (e.g., financial product-user-financial product) to find path instances connecting the first and second financial products in a heterogeneous target information network, count the number of such path instances, and include this count as part of the similarity calculation. Then, using a similarity metric suitable for heterogeneous information networks, combined with the number of path instances and other indicators, the similarity between the first financial product and each of the second financial products is calculated.

[0069] Step S204: Based on the influence of N first financial products and the similarity between each of the N first financial products and each second financial product, determine the target financial product from the second financial products and recommend the target financial product to the target user.

[0070] Optionally, the influence value of each first financial product can be used as a weight, combined with their similarity to the second financial products, and a weighted average or other aggregation strategy can be used to calculate the recommendation level value of each second financial product. Based on the recommendation level values ​​of the second financial products, the target financial product can be determined from among the second financial products. For example, the top M second financial products with the highest recommendation level values ​​can be identified as the target financial products. M is a positive integer.

[0071] In an optional embodiment, product information of the target financial product can be sent to the target user's device terminal via email, SMS, telephone, application software message reminder, etc., so as to recommend the target financial product to the target user.

[0072] In this embodiment, by constructing a target heterogeneous information network, the association between users and financial products is comprehensively considered. The influence of the first financial product associated with the target user is determined based on the target heterogeneous information network, and the similarity between the first financial product and other financial products (i.e., each second financial product) is determined. This realizes the comprehensive association relationship in the target heterogeneous information network, determines the importance of the first financial product that the user is concerned about and the degree of association between the first financial product and other financial products, thereby effectively improving the accuracy of product recommendation when determining the target financial product based on influence and similarity.

[0073] Therefore, the method provided in this application achieves the goal of calculating the influence and similarity of products based on the target heterogeneous information network, and making product recommendations based on influence and similarity, thereby improving the technical effect of improving the accuracy of product recommendations and solving the technical problem of low accuracy of product recommendations in related technologies.

[0074] Optionally, in the product recommendation method provided in this application embodiment, determining the influence of the first financial product based on the target heterogeneous information network includes: determining at least one network path conforming to the target meta-path from the target heterogeneous information network, wherein the target meta-path is a path template represented by node type, the node type of the start and end points of the target meta-path is the product type, and the network path is formed by connecting nodes in the target heterogeneous information network; determining the influence of the first financial product based on the number of products corresponding to the starting node in the at least one network path, the influence of the target neighbor node, and the number of nodes associated with the target neighbor node, wherein the target neighbor node is the neighbor node of the first financial product in the at least one network path.

[0075] Metapath refers to a path within a given network pattern Below, different object types and The conformity between It can be obtained from the following meta-paths This indicates that the metapath represents the path from object type in the network pattern. arrive A meta-path is a composite path that contains both object type and relation type. In other words, a meta-path is a template path.

[0076] In an optional embodiment, a target meta-path defines a path that starts from a product type node, passes through a series of specific type nodes, and reaches another product type node. For example, a target meta-path could be "product-user-product," meaning starting from a product, passing through user nodes, and finding related products.

[0077] Optionally, a network path refers to a specific path instance identified from the target heterogeneous information network based on the target meta-path. That is, the network path includes specific nodes in the target heterogeneous information network. For example, an optional network path could be "Financial Product A - User 1 - Financial Product B".

[0078] For example, the target processing system determines the target meta-path as "Product-User-Product" (PUP). This meta-path defines the path starting from a financial product, passing through a user node, and then reaching another financial product. Then, using a graph traversal algorithm or path search algorithm, all network path instances that conform to the above target meta-path are found in the network, resulting in at least one network path.

[0079] After determining at least one network path, the influence of the first financial product is determined based on the number of products corresponding to the starting node in the at least one network path, the influence of the target neighbor nodes, and the number of nodes associated with the target neighbor nodes. For example, the influence of the first financial product can be determined using the following formula:

[0080]

[0081] in, It refers to the first financial product. Q represents the influence of the first financial product. Q represents the number of products corresponding to the starting node in at least one network path. For example, if there are 5 network paths, and the starting node of 3 of them is financial product A, and the starting nodes of the other two network paths are financial products B and C respectively, then Q = 3. The damping coefficient is 0~1. Indicates the first financial product The directly associated objects in the network path, that is, the target neighbor nodes, for example, first determine the target network path starting from the first financial product, and then determine the neighbor nodes of the first financial product in the target network path as the target neighbor nodes. Representation Object (That is, the number of direct associations between the target neighbor node y and other objects under the target meta-path P, i.e., the number of objects) The out-degree in a target heterogeneous information network, for example, assuming the target meta-path object y is type, It means standing At the node, how many pointers does object y have under the target metapath P? The edge.

[0082] In an optional embodiment, influence can be initialized for each node in the target heterogeneous information network. Then, the influence of each node (including the first financial product) in the target heterogeneous information network can be calculated using the above influence calculation formula. Based on the calculated influence, the next round of influence calculation can be performed. This process is repeated until a preset number of iterations is reached or the change in the influence of the node is less than a preset range. At this point, the influence of the first financial product is determined.

[0083] In an alternative embodiment, in a given target heterogeneous information network, for the target meta-path This can be achieved using an adjoining matrix. Calculate the start one by one Type objects and endpoints The number of metapath instances between type objects, that is, the number of network paths that match the target metapath.

[0084] It should be noted that by introducing the concept of target meta-path and quantifying the number of products involved in the network path and the information of the nodes associated with the first financial product, this method not only considers direct user behavior, but also mines the indirect relationship between users and products through meta-path, thereby improving the accuracy of determining the influence of the first financial product.

[0085] Optionally, in the product recommendation method provided in this application embodiment, before determining the influence of the first financial product based on the target heterogeneous information network, the method further includes: obtaining multiple candidate meta-paths; determining the sum of the influence of product type nodes in the target heterogeneous information network based on the candidate meta-paths; and determining the target meta-path from the multiple candidate meta-paths based on the sum of the influence corresponding to the multiple candidate meta-paths.

[0086] In heterogeneous information networks, a meta-path defines a composite path from one type of node to another, composed of nodes of different types and the types of relationships between them. In this embodiment, multiple candidate meta-paths contain various possible path templates, all of which start from a product type node and reach another product type node through one or more intermediate node types.

[0087] In an optional embodiment, to further improve the efficiency and accuracy of similarity queries, a relatively superior meta-path can be selected from multiple candidate meta-paths as the target meta-path. That is, in the target heterogeneous information network, when a meta-path maximizes the sum of the object influence values ​​of nodes of the product type on that meta-path, that meta-path is determined to be relatively superior. The calculation formula is as follows:

[0088]

[0089] in, Let S represent the set of multiple candidate meta-paths, and let S be the set of product type nodes in the target heterogeneous information network. Let s be the influence of object s under metapath P. The candidate metapath with the largest sum of influence is determined as the target metapath.

[0090] Optionally, the target processing system can calculate the influence of product type nodes under a specific candidate path based on the aforementioned influence calculation formula, and obtain the sum of the influence of product type nodes based on a summation algorithm. In simpler terms, the influence of product type nodes at this stage is calculated based on the aforementioned method for calculating the influence of the first financial product, and therefore will not be elaborated upon here.

[0091] In another optional embodiment, the total influence of the first financial product in the target heterogeneous information network can be determined based on candidate meta-paths. Then, based on the total influence of the first financial products corresponding to multiple candidate meta-paths, the target meta-path can be determined from the multiple candidate meta-paths. In this way, the targeting of product recommendations to users can be effectively improved.

[0092] It should be noted that by introducing the concept of candidate meta-paths and analyzing their summed influence, the selection process for target meta-paths is further refined, improving the representativeness and effectiveness of the meta-paths used for influence assessment. In the context of heterogeneous information networks, this approach can more comprehensively capture the complex relationship between users and financial products, as well as the product's own influence, thereby further improving recommendation accuracy.

[0093] Optionally, in the product recommendation method provided in this application embodiment, calculating the similarity between the first financial product and each second financial product based on the target heterogeneous information network includes: for each second financial product, determining a network path from the target heterogeneous information network that starts with the first financial product and ends with the second financial product, and conforms to the target meta-path, to obtain a first network path; determining a network path from the target heterogeneous information network that starts with the first financial product and ends with the first financial product, and conforms to the target meta-path, to obtain a second network path; determining a network path from the target heterogeneous information network that starts with the second financial product and ends with the second financial product, and conforms to the target meta-path, to obtain a third network path; and determining the similarity between the first financial product and the second financial product based on the number of paths in the first network path, the number of paths in the second network path, and the number of paths in the third network path.

[0094] Optionally, a target metapath defines a path that starts from a product type node, passes through a series of specific type nodes, and reaches another product type node. For example, a target metapath could be "product-user-product," meaning starting from a product, passing through user nodes, and finding related products.

[0095] For example, a target processing system can calculate the similarity between a first financial product and a second financial product using the following formula:

[0096]

[0097] in, This represents the similarity between the first financial product x and the second financial product z. This refers to a network path instance that satisfies the target meta-path between the first financial product x and the second financial product z, also known as the first network path. For example, if user 1 purchases both the first financial product x and the second financial product z, a first network path is formed: "first financial product x - user 1 - second financial product z". This represents a network path instance where the first financial product x itself satisfies the target meta-path, which is also the second network path. For example, if user 1 buys the first financial product x, a second network path is formed: "first financial product x - user 1 - first financial product x". This represents a network path instance where the second financial product z satisfies the target meta-path, i.e., the third network path.

[0098] For example, assuming the first network path has 2 paths, the second network path has 2 paths, and the third network path has 3 paths, then PathSim(x, z) = (2 2) / (2+3)=0.8.

[0099] It should be noted that by combining the quantities of the three paths, a comprehensive assessment of the correlation between the first and second financial products is achieved. This not only considers the direct connection between the products, but also incorporates their respective self-strength and importance under specific relationship patterns, thereby improving the accuracy of the calculated similarity.

[0100] Optionally, in the product recommendation method provided in this application embodiment, determining a target financial product from the second financial products based on the influence of N first financial products and the similarity between each of the N first financial products and each of the second financial products includes: for each second financial product, determining the target similarity between the second financial product and the first financial products based on the similarity between the second financial product and the first financial products and the influence of the first financial products; summing the target similarities between the second financial product and each of the first financial products to obtain the recommendation degree value of the second financial product; and determining the target financial product from the second financial products based on the recommendation degree value of the second financial product.

[0101] Optionally, when calculating the similarity between a specific second financial product and a certain first financial product, a weight can be determined based on the influence of the first financial product, and then the target similarity between the second financial product and the first financial product can be determined based on the weight and the similarity between the second financial product and the first financial product.

[0102] After determining the target similarity between the second financial product and all first financial products, the target similarity between the second financial product and all first financial products is summed to obtain the recommendation value of the second financial product.

[0103] After obtaining the recommendation scores for all secondary financial products, the top M secondary financial products with the highest recommendation scores can be identified as target financial products.

[0104] It should be noted that by determining the target similarity through the similarity between the second financial product and the first financial product, as well as the influence of the first financial product, the recommendation process considers not only users' past behavioral preferences but also the product's actual influence and popularity. By determining the recommendation degree value based on the similarity between the second financial product and all first financial products, the limitations of single-point evaluation are avoided, thereby further improving the accuracy of product recommendations.

[0105] Optionally, in the product recommendation method provided in this application embodiment, determining the target similarity between the second financial product and the first financial product based on the similarity between the second financial product and the first financial product and the influence of the first financial product includes: for each first financial product, normalizing the influence of the first financial product based on the sum of the influences of N first financial products to obtain the target influence of the first financial product; calculating the product between the similarity between the second financial product and the first financial product and the target influence of the first financial product to obtain the target similarity.

[0106] In an optional embodiment, the sum of the influence of N first financial products can be calculated, and then the influence of a specific first financial product can be divided by the sum of the influence of the N first financial products to obtain the target influence of the specific first financial product.

[0107] After obtaining the target influence, the target influence is used as a weight to calculate the similarity between the second financial product and the first financial product, and the product of the target influence of the first financial product, to obtain the target similarity.

[0108] In an optional embodiment, the recommendation level value of the second financial product can be calculated using the following formula:

[0109]

[0110] in, This represents the recommendation level value of the second financial product z. This represents the i-th first financial product. This represents the target influence of the i-th first financial product.

[0111] It should be noted that normalization eliminates the magnitude difference in the original influence values, making the similarity calculation more objective and accurate, thereby further improving the accuracy of the calculated target similarity.

[0112] Optionally, in the product recommendation method provided in this application embodiment, obtaining N first financial products associated with the target user includes: determining the financial products already purchased by the target user as first financial products, and / or, determining the first financial products from the financial products based on the interaction information between the target user and the financial products.

[0113] In an optional embodiment, the target processing system can access the target user's historical transaction data to identify all purchased financial products. For example, it can retrieve purchase records associated with the target user ID from a database, extract product IDs from them, and regard the financial products corresponding to these product IDs as the first financial product.

[0114] In an optional embodiment, the aforementioned interaction information includes, but is not limited to, user behavior data such as browsing, consulting, evaluating, and paying attention to financial products. This information can more comprehensively represent the user's interest and inclination towards financial products. For example, the target processing system accesses user behavior logs, clickstream data, and feedback information in the financial institution's software, and processes this data using big data analysis tools, neural network models, regular expressions, etc., to identify the financial products that the target user is interested in based on user interaction information, thus obtaining the first financial product. For example, the target processing system has a preset interest score calculation formula. The interest score calculation formula can calculate the user's degree of interest in financial products based on the frequency of various types of user interaction behaviors. When the degree of interest is greater than a preset threshold, the corresponding financial product is determined as the first financial product.

[0115] It should be noted that identifying the primary financial product based on the user's purchase history and / or product interaction information improves the accuracy of the identified primary financial product.

[0116] In an alternative embodiment, Figure 5 This is a schematic diagram of the target processing system provided according to an embodiment of this application, such as... Figure 5 As shown, the target processing system may include a heterogeneous information network construction module, a product similarity analysis module, a user purchase history analysis module, a financial product recommendation module, and a visual product display module. The modules can communicate with each other through data interfaces to achieve low coupling between modules and high cohesion of system modules.

[0117] Optionally, the heterogeneous information network construction module is used to construct the target heterogeneous information network. The product similarity analysis module is used to calculate the similarity between the first financial product and the second financial product, as well as the influence of the first financial product, based on the target heterogeneous information network. The user purchase history analysis module is used to analyze the user's purchase history of financial products to determine the first financial product. The financial product recommendation module is used to determine the target financial product based on the calculated similarity and influence, and recommend it to the user. The visual product display module is used to provide users with a visual display of the target financial product, such as using charts and graphs.

[0118] Therefore, the method provided in this application achieves the goal of calculating the influence and similarity of products based on the target heterogeneous information network, and making product recommendations based on influence and similarity, thereby improving the technical effect of improving the accuracy of product recommendations and solving the technical problem of low accuracy of product recommendations in related technologies.

[0119] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0120] Example 2

[0121] This application also provides a product recommendation device. It should be noted that the product recommendation device of this application can be used to execute the product recommendation method provided in this application. The product recommendation device provided in this application is described below.

[0122] According to embodiments of this application, an apparatus for implementing the above-described product recommendation method is also provided, such as... Figure 6 As shown, the device includes:

[0123] The first acquisition module 601 is used to acquire N first financial products associated with the target user, where N is a positive integer greater than 1;

[0124] The first determining module 602 is used to determine the influence of each first financial product based on a target heterogeneous information network. The influence represents the importance of the first financial product. The target heterogeneous information network includes at least the following types of nodes: user type and product type. The edges between nodes are used to represent the association between nodes. The target heterogeneous information network includes the first financial product and the second financial product.

[0125] The first calculation module 603 is used to calculate the similarity between the first financial product and each of the second financial products based on the target heterogeneous information network;

[0126] The processing module 604 is used to determine the target financial product from the second financial products based on the influence of N first financial products and the similarity between each of the N first financial products and each second financial product, and to recommend the target financial product to the target user.

[0127] In this embodiment, by constructing a target heterogeneous information network, the association between users and financial products is comprehensively considered. The influence of the first financial product associated with the target user is determined based on the target heterogeneous information network, and the similarity between the first financial product and other financial products (i.e., each second financial product) is determined. This realizes the comprehensive association relationship in the target heterogeneous information network, determines the importance of the first financial product that the user is concerned about and the degree of association between the first financial product and other financial products, thereby effectively improving the accuracy of product recommendation when determining the target financial product based on influence and similarity.

[0128] Therefore, the method provided in this application achieves the goal of calculating the influence and similarity of products based on the target heterogeneous information network, and making product recommendations based on influence and similarity, thereby improving the technical effect of improving the accuracy of product recommendations and solving the technical problem of low accuracy of product recommendations in related technologies.

[0129] Optionally, in the product recommendation device provided in this application embodiment, the first determining module further includes: a first determining submodule, used to determine at least one network path conforming to the target meta-path from the target heterogeneous information network, wherein the target meta-path is a path template represented by node type, the node type of the starting point and the ending point of the target meta-path is the product type, and the network path is formed by connecting nodes in the target heterogeneous information network; a second determining submodule, used to determine the influence of the first financial product based on the number of products corresponding to the starting node in the at least one network path, the influence of the target neighbor node, and the number of nodes associated with the target neighbor node, wherein the target neighbor node is the neighbor node of the first financial product in the at least one network path.

[0130] Optionally, in the product recommendation device provided in the embodiments of this application, the product recommendation device further includes: a second acquisition module, used to acquire multiple candidate meta-paths; a second calculation module, used to determine the total influence of nodes of product type in the target heterogeneous information network based on the candidate meta-paths; and a second determination module, used to determine the target meta-path from the multiple candidate meta-paths based on the total influence corresponding to the multiple candidate meta-paths.

[0131] Optionally, in the product recommendation device provided in this application embodiment, the first calculation module further includes: a third determining submodule, used to determine, for each second financial product, a network path from the target heterogeneous information network that starts with the first financial product and ends with the second financial product, and conforms to the target meta-path, to obtain a first network path; a fourth determining submodule, used to determine, from the target heterogeneous information network, a network path that starts with the first financial product and ends with the first financial product, and conforms to the target meta-path, to obtain a second network path; a fifth determining submodule, used to determine, from the target heterogeneous information network, a network path that starts with the second financial product and ends with the second financial product, and conforms to the target meta-path, to obtain a third network path; and a sixth determining submodule, used to determine the similarity between the first financial product and the second financial product based on the number of paths in the first network path, the number of paths in the second network path, and the number of paths in the third network path.

[0132] Optionally, in the product recommendation device provided in this application embodiment, the processing module further includes: a seventh determining submodule, used to determine the target similarity between the second financial product and the first financial product for each second financial product based on the similarity between the second financial product and the first financial product and the influence of the first financial product; a calculation submodule, used to sum the target similarities between the second financial product and each first financial product to obtain the recommendation degree value of the second financial product; and an eighth determining submodule, used to determine the target financial product from the second financial products based on the recommendation degree value of the second financial product.

[0133] Optionally, in the product recommendation device provided in this application embodiment, the seventh determining submodule further includes: a processing unit, used to normalize the influence of each first financial product based on the sum of the influences of N first financial products to obtain the target influence of the first financial product; and a calculation unit, used to calculate the product between the similarity between the second financial product and the first financial product and the target influence of the first financial product to obtain the target similarity.

[0134] Optionally, in the product recommendation device provided in the embodiments of this application, the first acquisition module further includes: a ninth determining submodule, used to determine the financial products purchased by the target user as the first financial product, and / or a tenth determining submodule, used to determine the first financial product from the financial products based on the interaction information between the target user and the financial products.

[0135] It should be noted that the first acquisition module 601, the first determination module 602, the first calculation module 603, and the processing module 604 mentioned above correspond to steps S201 to S204 in Embodiment 1. The four modules and the corresponding steps implement the same instances and application scenarios, but are not limited to the content disclosed in Embodiment 1. It should be noted that the above modules or units can be hardware or software components stored in memory (e.g., memory 104) and processed by one or more processors (e.g., processors 102a, 102b, ..., 102n). The above modules can also be part of a device and run in the computer terminal 10 provided in Embodiment 1.

[0136] Example 3

[0137] Embodiments of this application may provide an electronic device. Figure 7 This is a structural block diagram of an electronic device according to an embodiment of this application. Figure 7 As shown, the electronic device may include: one or more ( Figure 7 (Only one is shown) processor 1002, memory 1004, memory controller, and peripheral interface, wherein the peripheral interface is connected to the radio frequency module, audio module and display.

[0138] The memory can be used to store software programs and modules, such as the program instructions / modules corresponding to the methods and apparatus in the embodiments of this application. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory, thereby implementing the above-described methods. The memory may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory may further include memory remotely located relative to the processor, and these remote memories can be connected to the terminal via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0139] The processor can access information and applications stored in memory via a transmission device to perform the following steps: Obtain N first financial products associated with the target user, where N is a positive integer greater than 1; for each first financial product, determine its influence based on a target heterogeneous information network, where influence represents the importance of the first financial product. The target heterogeneous information network includes at least the following types of nodes: user type, product type, and edges between nodes representing the relationships between nodes. The target heterogeneous information network includes first financial products and second financial products; calculate the similarity between the first financial product and each second financial product based on the target heterogeneous information network; based on the influence of the N first financial products and the similarity between the N first financial products and each second financial product, determine the target financial product from the second financial products and recommend the target financial product to the target user.

[0140] The processor can also invoke information and applications stored in the memory via a transmission device to perform the following steps: determining at least one network path conforming to the target meta-path from the target heterogeneous information network, wherein the target meta-path is a path template represented by node type, the node types of the start and end points of the target meta-path are product types, and the network path is formed by connecting nodes in the target heterogeneous information network; determining the influence of the first financial product based on the number of products corresponding to the starting node in the at least one network path, the influence of the target neighbor node, and the number of nodes associated with the target neighbor node, wherein the target neighbor node is the neighbor node of the first financial product in the at least one network path.

[0141] The processor can also invoke information and applications stored in the memory via a transmission device to perform the following steps: before determining the influence of the first financial product based on the target heterogeneous information network, obtain multiple candidate meta-paths; based on the candidate meta-paths, determine the sum of the influence of nodes of the product type in the target heterogeneous information network; based on the sum of the influence corresponding to the multiple candidate meta-paths, determine the target meta-path from the multiple candidate meta-paths.

[0142] The processor can also invoke information and applications stored in the memory via a transmission device to perform the following steps: For each second financial product, determine a network path from the target heterogeneous information network that starts with the first financial product, ends with the second financial product, and conforms to the target meta-path, to obtain a first network path; determine a network path from the target heterogeneous information network that starts with the first financial product and ends with the second financial product, and conforms to the target meta-path, to obtain a second network path; determine a network path from the target heterogeneous information network that starts with the second financial product and ends with the second financial product, and conforms to the target meta-path, to obtain a third network path; determine the similarity between the first financial product and the second financial product based on the number of paths in the first network path, the number of paths in the second network path, and the number of paths in the third network path.

[0143] The processor can also call the information and application programs stored in the memory through the transmission device to perform the following steps: for each second financial product, determine the target similarity between the second financial product and the first financial product based on the similarity between the second financial product and the first financial product and the influence of the first financial product; sum the target similarities between the second financial product and each first financial product to obtain the recommendation degree value of the second financial product; and determine the target financial product from the second financial products based on the recommendation degree value of the second financial product.

[0144] The processor can also call the information and application stored in the memory through the transmission device to perform the following steps: For each first financial product, normalize the influence of the first financial product based on the sum of the influence of N first financial products to obtain the target influence of the first financial product; calculate the product between the similarity between the second financial product and the first financial product and the target influence of the first financial product to obtain the target similarity.

[0145] The processor can also call the information and application stored in the memory through the transmission device to perform the following steps: identifying the financial products purchased by the target user as the first financial product, and / or, identifying the first financial product from the financial products based on the interaction information between the target user and the financial products.

[0146] Those skilled in the art will understand that Figure 7 The structure shown is for illustrative purposes only. Electronic devices can also be smartphones, tablets, handheld computers, mobile internet devices (MIDs), PADs, and other terminal devices. Figure 7 This does not limit the structure of the aforementioned electronic device. For example, electronic devices may also include components that are more... Figure 7 The more or fewer components shown (such as network interfaces, display devices, etc.), or having the same Figure 7The different configurations shown.

[0147] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing the hardware related to the terminal device. The program can be stored in a computer-readable storage medium, which may include: flash drive, read-only memory (ROM), random access memory (RAM), disk or optical disk, etc.

[0148] Example 4

[0149] Embodiments of this application also provide a storage medium. Optionally, in this embodiment, the storage medium can be used to store the program code executed by the product recommendation method provided in Embodiment 1.

[0150] Optionally, in this embodiment, the storage medium may be located in any computer terminal in a group of computer terminals in a computer network, or in any mobile terminal in a group of mobile terminals.

[0151] This application also provides a computer program product that, when executed on a data processing device, is adapted to perform recommended method steps of the product.

[0152] The sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0153] In the above embodiments of this application, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0154] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection of units or modules may be electrical or other forms.

[0155] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0156] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0157] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as a USB flash drive, read-only memory (ROM), random access memory (RAM), portable hard drive, magnetic disk, or optical disk.

[0158] The above description is only a preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A product recommendation method, characterized in that, include: Retrieve the N first financial products associated with the target user, where N is a positive integer greater than 1; For each first financial product, the influence of the first financial product is determined based on a target heterogeneous information network, wherein the influence characterizes the importance of the first financial product, and the target heterogeneous information network includes at least the following types of nodes: user type, product type, and the edges between nodes are used to characterize the association between nodes. The target heterogeneous information network includes the first financial product and the second financial product. The similarity between the first financial product and each of the second financial products is calculated based on the target heterogeneous information network. Based on the influence of the N first financial products and the similarity between the N first financial products and each second financial product, a target financial product is determined from the second financial products and recommended to the target user.

2. The method according to claim 1, characterized in that, The influence of the first financial product is determined based on the target heterogeneous information network, including: Determine at least one network path that conforms to the target meta-path from the target heterogeneous information network, wherein the target meta-path is a path template represented by node type, the node type of the start and end points of the target meta-path is product type, and the network path is formed by connecting the nodes in the target heterogeneous information network; The influence of the first financial product is determined based on the number of products corresponding to the starting node in the at least one network path, the influence of the target neighbor node, and the number of nodes associated with the target neighbor node, wherein the target neighbor node is a neighbor node of the first financial product in the at least one network path.

3. The method according to claim 2, characterized in that, Before determining the influence of the first financial product based on the target heterogeneous information network, the method further includes: Obtain multiple candidate meta paths; Based on the candidate meta paths, the total influence of nodes of product type in the target heterogeneous information network is determined; The target meta-path is determined from the multiple candidate meta-paths based on the sum of their influence.

4. The method according to claim 1, characterized in that, Calculating the similarity between the first financial product and each of the second financial products based on the target heterogeneous information network includes: For each second financial product, a first network path is obtained by determining a network path from the target heterogeneous information network that starts with the first financial product, ends with the second financial product, and conforms to the target meta-path. From the target heterogeneous information network, a network path that starts and ends with the first financial product and conforms to the target meta-path is determined to obtain the second network path; From the target heterogeneous information network, a network path that starts and ends with the second financial product and conforms to the target meta-path is determined to obtain a third network path; Based on the number of paths in the first network path, the number of paths in the second network path, and the number of paths in the third network path, the similarity between the first financial product and the second financial product is determined.

5. The method according to claim 1, characterized in that, Based on the influence of the N first financial products and the similarity between the N first financial products and each of the second financial products, target financial products are determined from the second financial products, including: For each second financial product, the target similarity between the second financial product and the first financial product is determined based on the similarity between the second financial product and the first financial product and the influence of the first financial product. The recommendation value of the second financial product is obtained by summing the target similarity between the second financial product and each of the first financial products. Based on the recommendation level value of the second financial product, the target financial product is determined from the second financial product.

6. The method according to claim 5, characterized in that, Based on the similarity between the second financial product and the first financial product, and the influence of the first financial product, the target similarity between the second financial product and the first financial product is determined, including: For each first financial product, the influence of the first financial product is normalized based on the sum of the influences of the N first financial products to obtain the target influence of the first financial product. The target similarity is obtained by multiplying the similarity between the second financial product and the first financial product by the target influence of the first financial product.

7. The method according to claim 1, characterized in that, Obtain N First Financial products associated with the target user, including: The financial products already purchased by the target user are identified as the first financial product, and / or, Based on the interaction information between the target user and the financial product, the first financial product is determined from the financial products.

8. A product recommendation device, characterized in that, include: The first acquisition module is used to acquire N first financial products associated with the target user, where N is a positive integer greater than 1; The first determining module is used to determine the influence of each first financial product based on a target heterogeneous information network, wherein the influence represents the importance of the first financial product, and the target heterogeneous information network includes at least the following types of nodes: user type, product type, and the edges between nodes are used to represent the association between nodes. The target heterogeneous information network includes the first financial product and the second financial product. The first calculation module is used to calculate the similarity between the first financial product and each of the second financial products based on the target heterogeneous information network; The processing module is used to determine a target financial product from the second financial products based on the influence of the N first financial products and the similarity between the N first financial products and each second financial product, and to recommend the target financial product to the target user.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored executable program, wherein, when the executable program is executed, it controls the device on which the computer-readable storage medium is located to perform the product recommendation method according to any one of claims 1 to 7.

10. An electronic device, characterized in that, include: Memory, which stores executable programs; A processor for running the program, wherein the program, when running, performs the product recommendation method according to any one of claims 1 to 7.

Images