Relationship graph-based information pushing method and device, electronic equipment and storage medium

Abstract

The application provides a relationship graph-based information pushing method and device, electronic equipment and storage medium. The method comprises the following steps: constructing a user relationship graph according to a pre-acquired product relationship between users, and constructing a product relationship graph according to a pre-acquired user relationship between products; the product relationship between users indicates whether two users have ever possessed the same product, and the user relationship between products indicates whether two products have ever been possessed by the same user; the user relationship graph comprises a plurality of user nodes, and the product relationship graph comprises a plurality of product nodes; the user node represents a user, and the product node represents a product; the connection probability between the user node and the product node is determined according to the connection relationship between the user nodes in the user relationship graph and the connection relationship between the product nodes in the product relationship graph; and a product is recommended to a user according to the connection probability between the user node and the product node. The effective information between the user and the product is fully mined, and the product recommendation accuracy is improved.

Description

Technical Field

[0001] This application relates to big data technology, and in particular to an information push method, device, electronic device and storage medium based on relationship graphs. Background Technology

[0002] With the continuous development of technology and the financial industry, major financial institutions have seized the opportunity presented by the mobile internet wave, launching a series of online platforms, and the variety and quantity of financial products have also grown rapidly. For financial institutions, the ability to utilize known customer information to recommend suitable financial products to each customer is crucial.

[0003] Currently, users are primarily categorized by analyzing their historical transaction data, and then suitable products are recommended based on rigid criteria. However, this approach lacks sufficient information mining of existing data, resulting in low recommendation accuracy and negatively impacting user experience. Summary of the Invention

[0004] This application provides a relationship graph-based information push method, apparatus, electronic device, and storage medium to improve the accuracy of product recommendations.

[0005] Firstly, this application provides an information push method based on a relationship graph, including:

[0006] Based on the pre-obtained product relationships between users, a user relationship graph is constructed, and based on the pre-obtained user relationships between products, a product relationship graph is constructed. The product relationships between users indicate whether two users have owned the same product, and the user relationships between products indicate whether two products have been owned by the same user. The user relationship graph includes at least two user nodes, and the product relationship graph includes at least two product nodes. One user node represents one user, and one product node represents one product.

[0007] Based on the connection relationships between user nodes in the user relationship graph and the connection relationships between product nodes in the product relationship graph, the connection probability between the user node and the product node is determined; wherein, the connection probability is used to represent the probability of recommending the product node's product to the user of the user node;

[0008] Based on the connection probability between the user node and the product node, the product information of the product to be pushed is determined, and the product information of the product to be pushed is pushed to the user terminal corresponding to the user node.

[0009] Secondly, this application provides an information push device based on a relationship graph, comprising:

[0010] The relationship graph construction module is used to construct a user relationship graph based on the product relationships between users obtained in advance, and to construct a product relationship graph based on the user relationships between products obtained in advance; wherein, the product relationships between users are used to indicate whether two users have owned the same product, and the user relationships between products are used to indicate whether two products have been owned by the same user; the user relationship graph includes at least two user nodes, and the product relationship graph includes at least two product nodes; one user node represents one user, and one product node represents one product;

[0011] The probability determination module is used to determine the connection probability between a user node and a product node based on the connection relationships between user nodes in the user relationship graph and the connection relationships between product nodes in the product relationship graph; wherein the connection probability is used to represent the probability of recommending the product node's product to the user of the user node;

[0012] The product recommendation module is used to determine the product information of the product to be pushed to the user terminal corresponding to the user node based on the connection probability between the user node and the product node.

[0013] Thirdly, this application provides an electronic device, including: a processor, and a memory communicatively connected to the processor;

[0014] The memory stores computer-executed instructions;

[0015] The processor executes computer execution instructions stored in the memory to implement the graph-based information push method as described in the first aspect of this application.

[0016] Fourthly, this application provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement the graph-based information push method as described in the first aspect of this application.

[0017] Fifthly, this application provides a computer program product, including a computer program that, when executed by a processor, implements the graph-based information push method described in the first aspect of this application.

[0018] This application provides an information push method, apparatus, electronic device, and storage medium based on relationship graphs. It constructs user relationship graphs and product relationship graphs by acquiring product relationships between users and user relationships between products. Based on the connections between user nodes in the user relationship graph and between product nodes in the product relationship graph, it determines the connection probability between user nodes and product nodes, i.e., the probability of recommending the product corresponding to the product node to the user corresponding to the user node, thereby pushing product information to the user. This solves the problem of low push accuracy caused by prior art that only pushes information to users based on historical transaction data. It fully explores the correlations between users, between products, and between users and products, improving the accuracy of information push. Attached Figure Description

[0019] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0020] Figure 1 A flowchart illustrating an information push method based on a relationship graph, provided in an embodiment of this application;

[0021] Figure 2 This is a schematic diagram of the user relationship provided in the embodiments of this application;

[0022] Figure 3 A flowchart illustrating an information push method based on a relationship graph, provided in an embodiment of this application;

[0023] Figure 4 A structural block diagram of an information push device based on a relationship graph provided in this application embodiment;

[0024] Figure 5 A structural block diagram of an information push device based on a relationship graph provided in this application embodiment;

[0025] Figure 6 A structural block diagram of an electronic device provided in an embodiment of this application;

[0026] Figure 7 This is a structural block diagram of an electronic device provided in an embodiment of this application.

[0027] The accompanying drawings have illustrated specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to specific embodiments. Detailed Implementation

[0028] To make the objectives, technical solutions, and advantages of this application clearer, the embodiments of this application will be described in further detail below with reference to the accompanying drawings.

[0029] It should be understood that the described embodiments are merely some, not all, of the embodiments in this application. All other embodiments obtained by those skilled in the art based on the embodiments in this application without inventive effort are within the scope of protection of this application.

[0030] In the following description, when referring to the accompanying drawings, the same numbers in different drawings denote the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0031] In the description of this application, it should be understood that the terms "first," "second," "third," etc., are used only to distinguish similar objects and are not necessarily used to describe a specific order or sequence, nor should they be construed as indicating or implying relative importance. Those skilled in the art can understand the specific meaning of the above terms in this application according to the specific circumstances. Furthermore, in the description of this application, unless otherwise stated, "multiple" refers to two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. The character " / " generally indicates that the preceding and following related objects have an "or" relationship.

[0032] It should be noted that, due to space limitations, this application specification does not exhaustively list all possible implementation methods. Those skilled in the art, after reading this application specification, should be able to deduce that, as long as the technical features do not contradict each other, any combination of technical features can constitute an optional implementation method. The following provides a detailed description of each embodiment.

[0033] It should be noted that the user information in this embodiment is not specific to any particular user and does not reflect the personal information of any particular user. The user information and product information in this embodiment come from publicly available datasets. The collection, storage, use, processing, transmission, provision, and disclosure of user personal information involved in this embodiment all comply with relevant laws and regulations and do not violate public order and good morals.

[0034] With the continuous development of technology and the financial industry, major financial institutions have seized the opportunity presented by the mobile internet wave, launching a series of online platforms. The variety and quantity of financial products have also grown rapidly, leading to increasingly fierce competition among financial institutions. This has placed higher demands on their work. For financial institutions, the ability to utilize known customer information to recommend suitable financial products to each customer is crucial.

[0035] Traditional methods primarily classify users by analyzing their historical transaction data and then use rigid criteria to select suitable products for recommendation. In recent years, methods such as random forests, logistic regression, and LSTM (Long Short-Term Memory) have been proposed to model user characteristics and historical transaction data, thereby predicting products that users might be interested in and recommending suitable products.

[0036] However, these methods ignore the correlations between users and between products, and do not sufficiently mine information from existing data, resulting in low accuracy of recommendations and affecting user experience.

[0037] This application provides an information push method, apparatus, electronic device, and storage medium based on a relationship graph, which aims to solve the above-mentioned technical problems in the prior art.

[0038] The technical solution of this application and how the technical solution of this application solves the above-mentioned technical problems are described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will be described below with reference to the accompanying drawings.

[0039] Figure 1 This is a flowchart illustrating an information push method based on a relationship graph, according to an embodiment of this application. This method can be executed by an information push device based on a relationship graph. Figure 1 As shown, the method includes the following steps:

[0040] S101. Based on the product relationships between users obtained in advance, construct a user relationship diagram, and based on the user relationships between products obtained in advance, construct a product relationship diagram.

[0041] For example, the product can be a financial product, and the user can be a customer who purchases the financial product. The product relationship between any two users and the user relationship between any two products are obtained in advance. For example, the purchase details of each user and the purchase status of each product can be collected in advance. The product relationship between users refers to the relationship between the products purchased by different users. The product relationship between users can be used to indicate whether two users have possessed the same product, that is, whether two users have purchased the same product. The product relationship between users can include whether two users have possessed the same product and whether two users have not possessed the same product. If two users have purchased the same product, that is, they have possessed the same product, then it is determined that there is a connection relationship between the two user nodes corresponding to these two users; if two users have not possessed the same product, then it is determined that there is no connection relationship between the user nodes corresponding to these two users. The more times two users have purchased the same product, the stronger the connection relationship between the two user nodes.

[0042] User relationships between products can be used to indicate whether two different products have been owned by the same user, i.e., whether two products have been purchased by the same user. User relationships between products can include whether two products have been owned by the same user and whether two products have not been owned by the same user. If two products have been purchased by the same user, then a connection exists between the two product nodes corresponding to these two products; if two products have not been owned by the same user, then no connection exists between the product nodes corresponding to these two products. The more users who purchase both types of products, the stronger the connection between the two product nodes. A user relationship graph includes at least two user nodes, and a product relationship graph includes at least two product nodes. A user node represents a user, and a product node represents a product.

[0043] Connecting two user nodes that have a connection relationship with an edge yields a user relationship graph. Connecting two product nodes that have a connection relationship with an edge yields a product relationship graph.

[0044] In this embodiment, a user relationship graph is constructed based on the product relationships between users obtained in advance, and a product relationship graph is constructed based on the user relationships between products obtained in advance. This includes: if two users have owned the same product, then an edge is established between the user nodes corresponding to the two users to obtain a user relationship graph; and if two products have been owned by the same user, then an edge is established between the product nodes corresponding to the two products to obtain a product relationship graph.

[0045] Specifically, if two users have owned the same product (i.e., purchased the same product), then a connection is established between their user nodes, and an edge can be created between them. If two users have not owned the same product, then no edge connection is needed between their user nodes. After determining whether an edge connection exists between any two users, a user relationship graph is obtained.

[0046] If two products have been owned by the same user (i.e., one user has purchased both products), then a connection exists between the product nodes of these two products, and an edge can be established between them. If two products have not been owned by the same user, then no edge connection is needed between their product nodes. After determining whether an edge connection exists between any two products, a product relationship graph is obtained.

[0047] The connections between nodes in a relationship graph can be represented using a matrix. If two nodes are connected, the value is 1; otherwise, the value is 0. Taking a user relationship graph as an example... Figure 2 As a schematic diagram of the user relationship, with three user nodes, a 3×3 matrix can be used. The matrix D1 representing the connection relationships between the user nodes can be expressed as:

[0048]

[0049] In this matrix, the elements on the main diagonal represent connections between identical user nodes, and are all 0. For example, the first row and first column represent the connection between user node 1 and user node 2. Elements outside the main diagonal can represent connections between two different user nodes. For example, a value of 1 in the third row and second column indicates a connection between user node 3 and user node 2. The connection matrix is ​​symmetric along the main diagonal, with elements at symmetrical positions being identical. For example, the first row and second column represent the connection between user node 1 and user node 2, and the second row and first column represent the connection between user node 2 and user node 1; these two connections are identical.

[0050] The benefits of this setup are that by determining the correlation between users and constructing a relationship graph between them, the relationships between users are fully explored, thus improving the accuracy of product recommendations; similarly, by determining the correlation between products and constructing a relationship graph between them, the relationships between products are fully explored, thus improving the accuracy of product recommendations.

[0051] S102. Based on the connection relationships between user nodes in the user relationship diagram and the connection relationships between product nodes in the product relationship diagram, determine the connection probability between user nodes and product nodes.

[0052] For example, determining the connection relationship between every two user nodes in the user relationship graph, i.e., determining whether there is an edge connection between every two user nodes. Similarly, determining the connection relationship between every two product nodes in the product relationship graph, i.e., determining whether there is an edge connection between every two product nodes.

[0053] Based on the connections between user nodes in the user relationship graph and between product nodes in the product relationship graph, the connection probability between user nodes and product nodes is determined; that is, the connection probability between two nodes of different types is determined. The connection probability can be used to represent the likelihood of pushing product information corresponding to a product node to a user of that user node. The higher the connection probability, the greater the likelihood of recommending the product corresponding to that product node to the user of that user node, i.e., the greater the likelihood of pushing product information to the user's terminal. For example, a user node is identified from the user relationship graph, and the products owned by the user corresponding to that user node are determined. The product node for that product is identified from the product relationship graph, and it is determined whether that product node has connections with other product nodes. If so, the product nodes connected to that product node are identified as associated nodes, and the connection probability between the user node and each associated node is relatively high. If a user node has an ownership relationship with a product node, then it is not necessary to determine the connection probability between that user node and that product node.

[0054] In this embodiment, a formula for calculating the connection probability can be preset, and the connection probability can be determined according to this formula. For example, the connection probability can be calculated based on the length of time a user occupies the product. In this embodiment, the formula for calculating the connection probability is not specifically limited. The connection probability between each user node and each product node can be determined.

[0055] S103. Based on the connection probability between the user node and the product node, determine the product information of the product to be pushed, and push the product information of the product to be pushed to the user terminal corresponding to the user node.

[0056] For example, after determining the connection probability between each user node and each product node, it can be determined whether to recommend the product corresponding to the product node to the user corresponding to each user node. For instance, a probability threshold can be preset. If the connection probability between a user node and a product node is equal to or greater than the preset probability threshold, then the product corresponding to that product node is recommended to the user corresponding to that user node; if the connection probability between a user node and a product node is less than the preset probability threshold, then the product corresponding to that product node is not recommended to the user corresponding to that user node. When recommending products to users, the products to be recommended can be determined, and product information can be obtained, such as the product name and product risk rate. This product information is then used as the product information of the product to be pushed, and the product information of the product to be pushed is sent to the user's bound terminal device to complete the information push.

[0057] A heterogeneous graph of users and products can be constructed based on the connection probability between user nodes and product nodes. This graph contains at least two different types of nodes. In this embodiment, the heterogeneous graph includes both user nodes and product nodes. If the connection probability between a user node and a product node is equal to or greater than a preset probability threshold, an edge connection is established between them; if the connection probability is less than the preset threshold, no edge connection is established. The connection relationships in the heterogeneous graph can also include connections between user nodes and connections between product nodes. After obtaining the heterogeneous graph, product recommendations can be made to users based on the connection relationships between user nodes and product nodes. For example, if a connection relationship exists between a user node and a product node, the product corresponding to the product node is recommended to the user corresponding to the user node. Product information for the recommended products can be retrieved from a preset database and pushed to the user's bound terminal device.

[0058] In this embodiment, the product information of the product to be pushed is determined based on the connection probability between the user node and the product node, and the product information of the product to be pushed is pushed to the user terminal corresponding to the user node. This includes: sorting the connection probabilities between the user node and each product node, determining the product node corresponding to the connection probability at the preset sorting position as the target product node; obtaining the product information of the product corresponding to the target product node as the product information of the product to be pushed, and pushing the product information of the product to be pushed to the user terminal corresponding to the user node.

[0059] Specifically, the connection probabilities between each user node and each product node are determined; that is, a user node can correspond to multiple connection probabilities. The connection probabilities of a user node are then sorted, for example, from largest to smallest. A pre-set sorting position is established, for example, the top ten positions. The product nodes corresponding to the connection probabilities at the pre-set sorting positions are then identified as target product nodes; for example, the product nodes corresponding to the top ten connection probabilities are designated as target product nodes.

[0060] The product information corresponding to the target product node is determined and used as the product information to be pushed to the user. This product information may include the product name, etc. The product information corresponding to the target product node is then pushed to the user corresponding to the user node; that is, the product corresponding to the target product node is recommended to the user. For example, product information can be pushed to the user's linked mobile terminal.

[0061] The advantage of this setup is that it uses connection probabilities to determine the potential connection relationship between users and products. A higher connection probability indicates a stronger correlation between the user node and the product node, suggesting a need to recommend products to the user. By identifying all connection probabilities for a given user node and selecting the top N probabilities with the highest values, products corresponding to these N probabilities are recommended to the user. This approach fully leverages the effective information within user-product transaction data, improving the accuracy and efficiency of information delivery.

[0062] This application provides an information push method based on relationship graphs. It constructs user relationship graphs and product relationship graphs by acquiring product relationships between users and user relationships between products. Based on the connections between user nodes in the user relationship graph and between product nodes in the product relationship graph, it determines the connection probability between user nodes and product nodes, i.e., the probability of recommending the product corresponding to the product node to the user corresponding to the user node, thereby pushing product information to the user. This solves the problem of low push accuracy caused by prior art that only pushes information to users based on their historical transaction data. It fully explores the correlations between users, between products, and between users and products, improving the accuracy of information push.

[0063] Figure 3 This is a flowchart illustrating an information push method based on a relationship graph, which is an optional embodiment based on the above embodiments.

[0064] In this embodiment, determining the connection probability between user nodes and product nodes based on the connection relationships between user nodes in the user relationship graph and the connection relationships between product nodes in the product relationship graph can be further refined as follows: obtaining user information corresponding to each user node in the user relationship graph and product information corresponding to each product node in the product relationship graph; determining user node representation information for each user node based on the connection relationships between user nodes in the user relationship graph and the user information corresponding to each user node in the user relationship graph, and determining product node representation information for each product node based on the connection relationships between product nodes in the product relationship graph and the product information corresponding to each product node in the product relationship graph; wherein, user node representation information is used to represent the user information and connection status of user nodes, and product node representation information is used to represent the product information and connection status of product nodes; and determining the connection probability between user nodes and product nodes based on the user node representation information and the product node representation information.

[0065] like Figure 3 The method includes the following steps:

[0066] S301. Based on the product relationships between users obtained in advance, construct a user relationship diagram, and based on the user relationships between products obtained in advance, construct a product relationship diagram.

[0067] For example, this step can refer to step S101 above, and will not be repeated here.

[0068] S302. Obtain the user information corresponding to each user node in the user relationship diagram, and the product information corresponding to each product node in the product relationship diagram.

[0069] For example, user information for each user and product information for each product are collected in advance. User information may include user name, age, gender, and assets, while product information may include product name, purchase amount, rate of return, and risk rate. The obtained user information is the user information corresponding to the user node in the user relationship graph; the obtained product information is the product information corresponding to the product node in the product relationship graph. User information of users who have made transactions and product information of products published by financial institutions can be obtained in real time or at regular intervals.

[0070] A user relationship graph can be constructed based on user information and the connections between user nodes. Each user node in the graph can include its user information. For example, if the user relationship graph is G1, its matrix can be represented as G1 = (X1, D1), where X1 represents the matrix of user information and D1 represents the matrix of connections between user nodes. Similarly, a product relationship graph can be constructed based on product information and the connections between product nodes. Each product node in the graph can include its product information. For example, if the product relationship graph is G2, its matrix can be represented as G2 = (X2, D2), where X2 represents the matrix of product information and D2 represents the matrix of connections between product nodes.

[0071] S303. Based on the connection relationships between user nodes in the user relationship diagram and the user information corresponding to each user node in the user relationship diagram, determine the user node representation information of each user node; and based on the connection relationships between product nodes in the product relationship diagram and the product information corresponding to each product node in the product relationship diagram, determine the product node representation information of each product node.

[0072] For example, feature extraction is performed on nodes in the user relationship graph and product relationship graph respectively to obtain user node representation information and product node representation information. Based on the connection relationships between user nodes and the user information of each user node in the user relationship graph, feature extraction is performed on each user node in the user relationship graph to determine the user node representation information. Based on the connection relationships between product nodes and the product information of each product node in the product relationship graph, feature extraction is performed on each product node in the product relationship graph to determine the product node representation information. User node representation information can be used to represent the user information and connection status of user nodes, and product node representation information can be used to represent the product information and connection status of product nodes. For example, a portion of the user information of a user node can be selected as user node representation information. Alternatively, the connection status of a user node can be used as user node representation information; for example, the number and location of neighboring nodes adjacent to a user node can be used as user node representation information.

[0073] In this embodiment, the user node representation information of each user node is determined based on the connection relationships between user nodes in the user relationship graph and the user information corresponding to each user node in the user relationship graph. This includes: determining the first neighbor node corresponding to each user node in the user relationship graph based on the connection relationships between user nodes in the user relationship graph; wherein, the first neighbor node is used to represent a node that has a connection relationship with the user node; and determining the user node representation information of the user node based on the user information of each first neighbor node corresponding to the user node.

[0074] Specifically, for each user node in the user relationship graph, based on the connections between user nodes in the graph, the neighboring nodes connected to each user node are determined as the first neighbor node corresponding to that user node. That is, the first neighbor node can represent a node that has a connection with the user node. A user node can have one or more first neighbor nodes. For example, Figure 2 The first neighbor nodes of user node 3 are user node 1 and user node 2.

[0075] The user information of each of the user node's first neighbor nodes is determined, and the user node representation information of the user node is determined based on the user information of each first neighbor node. For example, a preset portion of information can be selected from the user information of each first neighbor node, and the selected information can be summed or otherwise pre-calculated to extract the information features of the user node. In this embodiment, the method of calculating the user node representation information based on the user information of the first neighbor nodes is not specifically limited.

[0076] The advantage of this setup is that by extracting information features from user nodes through the connection relationships between them, new representation information of user nodes can be obtained, fully exploring the correlation between users, improving computational accuracy, and thus improving recommendation accuracy.

[0077] In this embodiment, the user node representation information of a user node is determined based on the user information of each of the first neighboring nodes corresponding to the user node. This includes: extracting user information features from the user node based on a preset graph neural network according to the user information of each of the first neighboring nodes corresponding to the user node, thereby obtaining the user node representation information of the user node.

[0078] Specifically, features of user nodes can be extracted using a pre-defined graph neural network (GNN) model. For example, the GNN could be a Graph Attention Network (GAT), a learning framework that aggregates information from neighboring nodes through an attention mechanism to obtain new node representations. The user information of each of the first neighboring nodes corresponding to a user node is input into the GNN model. The GNN model can include multiple convolutional layers to extract features from the user information. After feature extraction through these convolutional layers, the user node representation information is output.

[0079] For each convolutional layer in a graph neural network, features can be extracted from the information it receives using a pre-defined formula. The feature extraction formula can be as follows:

[0080]

[0081] in, This represents the data output by the u-th user node in the i-th convolutional layer. If the i-th convolutional layer is the last layer, then... This represents the user node information for the u-th user node. σ is a preset activation function, and N... u Let α be the set of the first neighbor nodes of the u-th user node, and v be the first neighbor nodes of user node u. uv The attention coefficients for the user node u and its first neighbor node v are preset; these coefficients can be between 0 and 1. W is a preset learnable parameter. This is the data of the first neighbor node v output from the (i-1)th layer. If i is 1, then... It contains the user information of the first neighbor node v.

[0082] The advantage of this setup is that by using a graph neural network to update the features of each user node in the user relationship graph, user node representation information is obtained, the correlation between users is fully explored, the computational accuracy and efficiency are improved, and thus the accuracy and efficiency of product recommendations are improved.

[0083] In this embodiment, the product node representation information of each product node is determined based on the connection relationship between product nodes in the product relationship diagram and the product information corresponding to each product node in the product relationship diagram. This includes: determining the second neighbor node corresponding to each product node in the product relationship diagram based on the connection relationship between product nodes in the product relationship diagram; wherein, the second neighbor node is used to represent a node that has a connection relationship with the product node; and determining the product node representation information of the product node based on the product information of each second neighbor node corresponding to the product node.

[0084] Specifically, for each product node in the product relationship graph, based on the connections between product nodes in the graph, the neighboring nodes connected to each product node are determined as the second neighboring nodes of that product node. That is, the second neighboring node can represent a node that has a connection to a product node. A product node can have one or more second neighboring nodes.

[0085] The product information of each of the second neighbor nodes of the product node is determined, and the product node representation information of the product node is determined based on the product information of each second neighbor node. For example, a preset portion of information can be selected from the product information of each second neighbor node, and the selected information can be summed or otherwise pre-calculated to extract the information features of the product node. In this embodiment, the method of calculating the product node representation information based on the product information of the second neighbor nodes is not specifically limited.

[0086] The beneficial effect of this setup is that by extracting information features from product nodes through the connection relationships between them, new representation information of product nodes can be obtained, the correlation between products can be fully explored, the calculation accuracy can be improved, and thus the recommendation accuracy can be improved.

[0087] In this embodiment, the product node representation information of a product node is determined based on the product information of each of the second neighboring nodes corresponding to the product node. This includes: extracting product information features from the product node based on a preset graph neural network according to the product information of each of the second neighboring nodes corresponding to the product node, thereby obtaining the product node representation information of the product node.

[0088] Specifically, features of product nodes can be extracted using a pre-defined graph neural network (GNN) model, such as a Gaussian graph neural network (GNN). The product information of each of the second neighbor nodes corresponding to a product node is input into the GNN model. The GNN model can include multiple convolutional layers to extract features from the product information. After feature extraction through multiple convolutional layers, the product node representation information is output.

[0089] For each convolutional layer in a graph neural network, features can be extracted from the information it receives using a pre-defined formula. The feature extraction formula can be as follows:

[0090]

[0091] in, This represents the data output by the k-th product node in the i-th convolutional layer. If the i-th convolutional layer is the last layer, then... This represents the product node information for the k-th product node. σ is a preset activation function, and N... k Let r be the set of the second neighbor nodes of the k-th product node, where r is the set of the second neighbor nodes of product node k. kr The attention coefficient is a preset value for the product node k and its second neighbor node r. The attention coefficient can be between 0 and 1. W is a preset learnable parameter. This is the data of the second neighbor node r output from the (i-1)th layer. If i is 1, then... This is the product information of the second neighbor node r.

[0092] The beneficial effect of this setup is that by using a graph neural network to update the features of each product node in the product relationship graph, product node representation information is obtained, the correlation between products is fully explored, the computational accuracy and efficiency are improved, and thus the accuracy and efficiency of product recommendations are improved.

[0093] S304. Determine the connection probability between the user node and the product node based on the user node representation information and the product node representation information.

[0094] For example, after obtaining user node representation information and product node representation information, calculations are performed on the user node representation information and product node representation information, and the calculated results are determined as the connection probability between the user node corresponding to the user node representation information and the product node corresponding to the product node representation information. For example, the weights of user node representation information and product node representation information can be preset, and the connection probability is calculated based on the preset weights. Alternatively, the connection probability can be calculated based on the user's ownership relationship with all products; the ownership relationship between the user and the product can be whether the user has purchased the product.

[0095] In this embodiment, determining the connection probability between a user node and a product node based on user node representation information and product node representation information includes: obtaining the ownership relationship between the user corresponding to the user node and each product; and determining the connection probability between the user node and the product node based on a preset probability prediction function, according to the user node representation information, product node representation information, and the ownership relationship between the user corresponding to the user node and each product.

[0096] Specifically, a probability prediction function is pre-defined. This function calculates the connection probability between a user node and a product node based on user node representation information, product node representation information, and the ownership relationship between the user corresponding to the user node and each product. The ownership relationship between the user corresponding to the user node and each product can be represented by whether the user has owned a product; if yes, it can be represented by 1; otherwise, it can be represented by 0. In this embodiment, the probability prediction function is not specifically limited. For example, the probability prediction function can be the DistMult factorization function. DistMult is a simple and effective factorization function that performs well when used alone as a probability prediction function.

[0097] The system obtains the ownership relationships between users and various products. It then substitutes the user node representation information, product node representation information, and the ownership relationships between users and products corresponding to each user node into a preset probability prediction function to calculate the connection probability between the user node corresponding to the user node representation information and the product node corresponding to the product node representation information. In this embodiment, the probability prediction function model can be trained using negative sampling, and the model can be optimized using a cross-entropy loss function.

[0098] The beneficial effect of this setup is that it mathematically models the user node representation information, product node representation information, and the ownership relationship between users and products, and uses a probability prediction function to make product recommendations, providing a new approach to product recommendation methods and improving the accuracy and efficiency of product recommendations.

[0099] S305. Based on the connection probability between the user node and the product node, determine the product information of the product to be pushed, and push the product information of the product to be pushed to the user terminal corresponding to the user node.

[0100] For example, this step can refer to step S103 above, and will not be repeated here.

[0101] This application provides an information push method based on relationship graphs. It constructs user relationship graphs and product relationship graphs by acquiring product relationships between users and user relationships between products. Based on the connections between user nodes in the user relationship graph and between product nodes in the product relationship graph, it determines the connection probability between user nodes and product nodes, i.e., the probability of recommending the product corresponding to the product node to the user corresponding to the user node, thereby pushing product information to the user. This solves the problem of low push accuracy caused by prior art that only pushes information to users based on their historical transaction data. It fully explores the correlations between users, between products, and between users and products, improving the accuracy of information push.

[0102] Figure 4 This is a structural block diagram of an information push device based on a relationship graph, provided as an embodiment of this application. For ease of explanation, only the parts related to the embodiments of this disclosure are shown. (Refer to...) Figure 4 The device includes: a relationship graph construction module 401, a probability determination module 402, and a product recommendation module 403.

[0103] The relationship graph construction module 401 is used to construct a user relationship graph based on the product relationships between users obtained in advance, and to construct a product relationship graph based on the user relationships between products obtained in advance; wherein, the product relationships between users are used to indicate whether two users have owned the same product, and the user relationships between products are used to indicate whether two products have been owned by the same user; the user relationship graph includes at least two user nodes, and the product relationship graph includes at least two product nodes; one user node represents one user, and one product node represents one product;

[0104] The probability determination module 402 is used to determine the connection probability between a user node and a product node based on the connection relationships between user nodes in the user relationship graph and the connection relationships between product nodes in the product relationship graph; wherein the connection probability is used to represent the probability of recommending the product of the product node to the user of the user node;

[0105] The product recommendation module 403 is used to determine the product information of the product to be pushed based on the connection probability between the user node and the product node, and push the product information of the product to be pushed to the user terminal corresponding to the user node.

[0106] Figure 5 This application provides a structural block diagram of an information push device based on a relationship graph, in which... Figure 4 Based on the illustrated embodiments, as Figure 5 As shown, the probability determination module 402 includes an information acquisition unit 4021, an information determination unit 4022, and a probability determination unit 4023.

[0107] The information acquisition unit 4021 is used to acquire user information corresponding to each user node in the user relationship diagram and product information corresponding to each product node in the product relationship diagram.

[0108] The information determination unit 4022 is used to determine user node representation information of each user node based on the connection relationships between user nodes in the user relationship diagram and the user information corresponding to each user node in the user relationship diagram, and to determine product node representation information of each product node based on the connection relationships between product nodes in the product relationship diagram and the product information corresponding to each product node in the product relationship diagram; wherein, the user node representation information is used to represent the user information and connection status of the user node, and the product node representation information is used to represent the product information and connection status of the product node;

[0109] The probability determination unit 4023 is used to determine the connection probability between the user node and the product node based on the user node representation information and the product node representation information.

[0110] In one example, the information determination unit 4022 includes:

[0111] The first node determination subunit is used to determine the first neighbor node corresponding to each user node in the user relationship graph based on the connection relationship between user nodes in the user relationship graph; wherein, the first neighbor node is used to represent a node that has a connection relationship with the user node;

[0112] The first information determination subunit is used to determine the user node representation information of the user node based on the user information of each of the first neighboring nodes corresponding to the user node.

[0113] In one example, the first piece of information identifies the sub-unit, specifically used for:

[0114] Based on the user information of each first neighbor node corresponding to the user node, user information features are extracted from the user node using a preset graph neural network to obtain the user node representation information of the user node.

[0115] In one example, the information determination unit 4022 includes:

[0116] The second node determining subunit is used to determine the second neighbor node corresponding to each product node in the product relationship diagram based on the connection relationship between product nodes in the product relationship diagram; wherein, the second neighbor node is used to represent a node that has a connection relationship with the product node;

[0117] The second information determination subunit is used to determine the product node representation information of the product node based on the product information of each of the second neighboring nodes corresponding to the product node.

[0118] In one example, the second information identifies the sub-unit, specifically used for:

[0119] Based on the product information of each second neighbor node corresponding to the product node, and using a preset graph neural network, product information features are extracted from the product node to obtain the product node representation information of the product node.

[0120] In one example, probability determination unit 4023 is specifically used for:

[0121] Obtain the ownership relationship between the user corresponding to the user node and each product;

[0122] Based on the user node representation information and the product node representation information, as well as the ownership relationship between the user corresponding to the user node and each product, the connection probability between the user node and the product node is determined based on a preset probability prediction function.

[0123] In one example, the product recommendation module 403 is specifically used for:

[0124] The connection probabilities between the user nodes and each product node are sorted, and the product node corresponding to the connection probability located at the preset sorting position is determined as the target product node.

[0125] Obtain the product information of the product corresponding to the target product node, which is the product information of the product to be pushed, and push the product information of the product to be pushed to the user terminal corresponding to the user node.

[0126] In one example, the relationship graph building module 401 includes:

[0127] The first construction unit is used to establish an edge between the user nodes corresponding to the two users if the two users have owned the same product, thereby obtaining a user relationship graph; and,

[0128] The second construction unit is used to establish an edge between the product nodes corresponding to the two products if the two products have been occupied by the same user, so as to obtain a product relationship graph.

[0129] Figure 6 A structural block diagram of an electronic device provided in an embodiment of this application, such as... Figure 6 As shown, the electronic device includes: a memory 61 and a processor 62; the memory 61 is a memory for storing executable instructions of the processor 62.

[0130] The processor 62 is configured to perform the methods provided in the above embodiments.

[0131] The electronic device also includes a receiver 63 and a transmitter 64. The receiver 63 is used to receive instructions and data sent by other devices, and the transmitter 64 is used to send instructions and data to external devices.

[0132] Figure 7 This is a structural block diagram of an electronic device according to an exemplary embodiment. The device may be a mobile phone, computer, digital broadcasting terminal, messaging device, game console, tablet device, medical device, fitness equipment, personal digital assistant, etc.

[0133] Device 700 may include one or more of the following components: processing component 702, memory 704, power supply component 706, multimedia component 708, audio component 710, input / output (I / O) interface 712, sensor component 714, and communication component 716.

[0134] Processing component 702 typically controls the overall operation of device 700, such as operations associated with display, telephone calls, data communication, camera operation, and recording. Processing component 702 may include one or more processors 720 to execute instructions to perform all or part of the steps of the methods described above. Furthermore, processing component 702 may include one or more modules to facilitate interaction between processing component 702 and other components. For example, processing component 702 may include a multimedia module to facilitate interaction between multimedia component 708 and processing component 702.

[0135] Memory 704 is configured to store various types of data to support the operation of device 700. Examples of this data include instructions for any application or method operating on device 700, contact data, phonebook data, messages, pictures, videos, etc. Memory 704 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0136] Power supply component 706 provides power to various components of device 700. Power supply component 706 may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power to device 700.

[0137] Multimedia component 708 includes a screen that provides an output interface between the device 700 and the user. In some embodiments, the screen may include a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen may be implemented as a touchscreen to receive input signals from the user. The touch panel includes one or more touch sensors to sense touches, swipes, and gestures on the touch panel. The touch sensors may sense not only the boundaries of the touch or swipe action but also the duration and pressure associated with the touch or swipe operation. In some embodiments, multimedia component 708 includes a front-facing camera and / or a rear-facing camera. When the device 700 is in an operating mode, such as a shooting mode or a video mode, the front-facing camera and / or the rear-facing camera may receive external multimedia data. Each front-facing camera and rear-facing camera may be a fixed optical lens system or have focal length and optical zoom capabilities.

[0138] Audio component 710 is configured to output and / or input audio signals. For example, audio component 710 includes a microphone (MIC) configured to receive external audio signals when device 700 is in an operating mode, such as call mode, recording mode, and voice recognition mode. The received audio signals may be further stored in memory 704 or transmitted via communication component 716. In some embodiments, audio component 710 also includes a speaker for outputting audio signals.

[0139] I / O interface 712 provides an interface between processing component 702 and peripheral interface modules, such as keyboards, click wheels, buttons, etc. These buttons may include, but are not limited to, home buttons, volume buttons, power buttons, and lock buttons.

[0140] Sensor assembly 714 includes one or more sensors for providing state assessments of various aspects of device 700. For example, sensor assembly 714 may detect the on / off state of device 700, the relative positioning of components such as the display and keypad of device 700, changes in the position of device 700 or a component of device 700, the presence or absence of user contact with device 700, the orientation or acceleration / deceleration of device 700, and temperature changes of device 700. Sensor assembly 714 may include a proximity sensor configured to detect the presence of nearby objects without any physical contact. Sensor assembly 714 may also include a light sensor, such as a CMOS or CCD image sensor, for use in imaging applications. In some embodiments, sensor assembly 714 may also include an accelerometer, a gyroscope, a magnetometer, a pressure sensor, or a temperature sensor.

[0141] Communication component 716 is configured to facilitate wired or wireless communication between device 700 and other devices. Device 700 can access wireless networks based on communication standards, such as WiFi, 2G, or 3G, or combinations thereof. In one exemplary embodiment, communication component 716 receives broadcast signals or broadcast-related information from an external broadcast management system via a broadcast channel. In one exemplary embodiment, communication component 716 also includes a near-field communication (NFC) module to facilitate short-range communication. For example, the NFC module may be implemented based on radio frequency identification (RFID) technology, Infrared Data Association (IrDA) technology, ultra-wideband (UWB) technology, Bluetooth (BT) technology, and other technologies.

[0142] In an exemplary embodiment, device 700 may be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field-programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to perform the methods described above.

[0143] In an exemplary embodiment, a non-transitory computer-readable storage medium including instructions is also provided, such as a memory 704 including instructions, which can be executed by a processor 720 of device 700 to perform the above-described method. For example, the non-transitory computer-readable storage medium may be a ROM, random access memory (RAM), CD-ROM, magnetic tape, floppy disk, and optical data storage device, etc.

[0144] A non-transitory computer-readable storage medium, when the instructions in the storage medium are executed by the processor of a terminal device, enables the terminal device to execute the graph-based information push method of the aforementioned device.

[0145] This application also discloses a computer program product, including a computer program that, when executed by a processor, implements the method described in this embodiment.

[0146] Various embodiments of the systems and technologies described above in this application can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include: implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0147] The program code used to implement the methods of this application may be written in any combination of one or more programming languages. This program code may be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing device, such that when executed by the processor or controller, the functions / operations specified in the flowcharts and / or block diagrams are implemented. The program code may be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or electronic device.

[0148] In the context of this application, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. Machine-readable media can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0149] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device for displaying information to the user (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor); and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the computer. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0150] The systems and technologies described herein can be implemented in computing systems that include back-end components (e.g., as data electronic devices), or computing systems that include middleware components (e.g., application electronic devices), or computing systems that include front-end components (e.g., a user computer with a graphical user interface or web browser through which a user can interact with implementations of the systems and technologies described herein), or any combination of such back-end, middleware, or front-end components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., a communication network). Examples of communication networks include local area networks (LANs), wide area networks (WANs), and the Internet.

[0151] Computer systems can include client and electronic devices. Clients and electronic devices are generally geographically separated and typically interact via communication networks. The client-electronic device relationship is created by computer programs running on the respective computers and having a client-electronic device relationship with each other. The electronic device can be a cloud electronic device, also known as a cloud computing electronic device or cloud host, a host product within the cloud computing service system, addressing the shortcomings of traditional physical hosts and VPS services ("Virtual Private Server," or simply "VPS") in terms of management difficulty and weak business scalability. The electronic device can also be an electronic device in a distributed system or an electronic device incorporating blockchain technology. It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this application can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution disclosed in this application is achieved, and this is not limited herein.

[0152] Other embodiments of this application will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of this application that follow the general principles of this application and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of this application are indicated by the following claims.

[0153] It should be understood that this application is not limited to the precise structure described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this application is limited only by the appended claims.

Claims

1. An information push method based on a relationship graph, characterized in that, include: Based on pre-obtained product relationships between users, a user relationship graph is constructed, and based on pre-obtained user relationships between products, a product relationship graph is constructed. The product relationships between users indicate whether two users have owned the same product, and the user relationships between products indicate whether two products have been owned by the same user. The user relationship graph includes at least two user nodes, and the product relationship graph includes at least two product nodes. One user node represents one user, and one product node represents one product. Constructing the user relationship graph and the product relationship graph based on pre-obtained user relationships includes: if two users have owned the same product, establishing an edge between the user nodes corresponding to the two users to obtain the user relationship graph; and if two products have been owned by the same user, establishing an edge between the product nodes corresponding to the two products to obtain the product relationship graph. The connection relationships between user nodes in the user relationship graph and the connection relationships between product nodes in the product relationship graph are represented in matrix form. If there is a connection between two nodes, the corresponding matrix element is 1, otherwise it is 0. The main diagonal elements of the matrix are all 0, and the matrix is ​​symmetric along the main diagonal. The process involves: acquiring user information corresponding to each user node in the user relationship graph and product information corresponding to each product node in the product relationship graph; determining the first neighbor node corresponding to each user node in the user relationship graph based on the connection relationships between user nodes in the user relationship graph; wherein the first neighbor node represents a node connected to the user node; determining user node representation information for each user node based on the user information of each first neighbor node corresponding to the user node; determining the second neighbor node corresponding to each product node in the product relationship graph based on the connection relationships between product nodes in the product relationship graph; wherein the second neighbor node represents a node connected to the product node; determining product node representation information for each second neighbor node corresponding to the product node based on the product information of each second neighbor node corresponding to the product node; wherein the user node representation information represents the user information and connection status of the user node, and the product node representation information represents the product information and connection status of the product node; and determining the connection probability between the user node and the product node based on the user node representation information and the product node representation information; wherein the connection probability represents the probability of recommending the product node's product to the user of the user node. Based on the connection probability between the user node and the product node, the product information of the product to be pushed is determined, and the product information of the product to be pushed is pushed to the user terminal corresponding to the user node.

2. The method according to claim 1, characterized in that, Based on the user information of each of the first neighboring nodes corresponding to the user node, the user node representation information of the user node is determined, including: Based on the user information of each first neighbor node corresponding to the user node, user information features are extracted from the user node using a preset graph neural network to obtain the user node representation information of the user node.

3. The method according to claim 1, characterized in that, Based on the product information of each of the second neighboring nodes corresponding to the product node, the product node representation information of the product node is determined, including: Based on the product information of each second neighbor node corresponding to the product node, and using a preset graph neural network, product information features are extracted from the product node to obtain the product node representation information of the product node.

4. The method according to claim 1, characterized in that, Determining the connection probability between the user node and the product node based on the user node representation information and the product node representation information includes: Obtain the ownership relationship between the user corresponding to the user node and each product; Based on the user node representation information and the product node representation information, as well as the ownership relationship between the user corresponding to the user node and each product, the connection probability between the user node and the product node is determined based on a preset probability prediction function.

5. The method according to any one of claims 1-4, characterized in that, Based on the connection probability between the user node and the product node, determine the product information of the product to be pushed, and push the product information of the product to be pushed to the user terminal corresponding to the user node, including: The connection probabilities between the user nodes and each product node are sorted, and the product node corresponding to the connection probability located at the preset sorting position is determined as the target product node. Obtain the product information of the product corresponding to the target product node, which is the product information of the product to be pushed, and push the product information of the product to be pushed to the user terminal corresponding to the user node.

6. An information push device based on a relationship graph, characterized in that, include: The relationship graph construction module is used to construct a user relationship graph based on the product relationships between users obtained in advance, and to construct a product relationship graph based on the user relationships between products obtained in advance; wherein, the product relationships between users are used to indicate whether two users have owned the same product, and the user relationships between products are used to indicate whether two products have been owned by the same user; the user relationship graph includes at least two user nodes, and the product relationship graph includes at least two product nodes; one user node represents one user, and one product node represents one product; The matrix representation module is used to represent the connection relationships between user nodes in the user relationship diagram and the connection relationships between product nodes in the product relationship diagram in matrix form; wherein, if there is a connection relationship between two nodes, the corresponding matrix element is 1, otherwise it is 0; the main diagonal elements of the matrix are all 0, and the matrix is ​​a symmetric matrix along the main diagonal; The probability determination module is used to determine the connection probability between a user node and a product node based on the connection relationships between user nodes in the user relationship graph and the connection relationships between product nodes in the product relationship graph; wherein the connection probability is used to represent the probability of recommending the product node's product to the user of the user node; The product recommendation module is used to determine the product information of the product to be pushed to the user terminal corresponding to the user node based on the connection probability between the user node and the product node. The relationship graph construction module includes: The first construction unit is used to establish an edge between the user nodes corresponding to the two users if the two users have owned the same product, thereby obtaining a user relationship graph; and, The second construction unit is used to establish an edge between the product nodes corresponding to the two products if the two products have been occupied by the same user, so as to obtain a product relationship graph. The probability determination module includes: The information acquisition unit is used to acquire user information corresponding to each user node in the user relationship diagram and product information corresponding to each product node in the product relationship diagram. The information determination unit is configured to determine user node representation information for each user node based on the connection relationships between user nodes in the user relationship diagram and the user information corresponding to each user node in the user relationship diagram, and to determine product node representation information for each product node based on the connection relationships between product nodes in the product relationship diagram and the product information corresponding to each product node in the product relationship diagram; wherein, the user node representation information is used to represent the user information and connection status of the user node, and the product node representation information is used to represent the product information and connection status of the product node; The probability determination unit is used to determine the connection probability between the user node and the product node based on the user node representation information and the product node representation information. The information determination unit includes: The first node determination subunit is used to determine the first neighbor node corresponding to each user node in the user relationship graph based on the connection relationship between user nodes in the user relationship graph; wherein, the first neighbor node is used to represent a node that has a connection relationship with the user node; The first information determination subunit is used to determine the user node representation information of the user node based on the user information of each of the first neighboring nodes corresponding to the user node. The information determination unit further includes: The second node determining subunit is used to determine the second neighbor node corresponding to each product node in the product relationship diagram based on the connection relationship between product nodes in the product relationship diagram; wherein, the second neighbor node is used to represent a node that has a connection relationship with the product node; The second information determination subunit is used to determine the product node representation information of the product node based on the product information of each of the second neighboring nodes corresponding to the product node.

7. An electronic device, characterized in that, include: A processor, and a memory communicatively connected to the processor; The memory stores computer-executed instructions; The processor executes the computer execution instructions stored in the memory to implement the information push method based on the relationship graph as described in any one of claims 1-5.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the graph-based information push method as described in any one of claims 1-5.

9. A computer program product, characterized in that, It includes a computer program that, when executed by a processor, implements the graph-based information push method according to any one of claims 1-5.

Images