Product recommendation method and product recommendation model training method

Abstract

Embodiments of the invention disclose a product recommendation method and a product recommendation model training method. The method comprises the steps of obtaining a target user feature matrix based on attribute feature data of a target user; inputting the target user feature matrix into a pre-trained product recommendation model, wherein the product recommendation model comprises an auto-encoder and a graph convolutional neural network; obtaining the attribute features of the target user based on the target user feature matrix through the auto-encoder to obtain the attribute features of the target user; and based on the target user attribute features, determining whether to recommend candidate products to the target user through the graph convolutional neural network to obtain a recommended product determination result. According to the embodiment of the invention, the recommendation accuracy and recommendation efficiency when the product is recommended to the target user can be improved.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence, and in particular to a product recommendation method and a product recommendation model training method. Background Art

[0002] With the continuous development of the economy, various products have emerged in an endless stream, including financial products and application products. How to accurately recommend suitable products to users and better provide personalized services to users has become a key issue for product providers. The product recommendation methods in the prior art include personalized recommendation methods based on matrix decomposition algorithms and collaborative filtering algorithms, but these methods only consider direct relationships in the recommendation process, and the recommendation accuracy is poor; in addition, due to classic recommendation algorithms such as matrix decomposition, the interaction matrix is ​​split into two feature matrices of users and items. In the case of large data sets, the time complexity of the split is too high and the computational efficiency is very low. Summary of the invention

[0003] The embodiments of the present invention provide a product recommendation method and a product recommendation model training method, which can improve the recommendation accuracy and recommendation efficiency when recommending products to users.

[0004] In a first aspect, an embodiment of the present invention provides a product recommendation method, comprising:

[0005] Acquire a target user feature matrix based on the attribute feature data of the target user;

[0006] Inputting the target user feature matrix into a pre-trained product recommendation model, wherein the product recommendation model includes an autoencoder and a graph convolutional neural network;

[0007] Obtaining the attribute features of the target user based on the target user feature matrix by the autoencoder to obtain the target user attribute features; and

[0008] Based on the attribute characteristics of the target user, the graph convolutional neural network is used to determine whether to recommend a candidate product to the target user to obtain a recommended product determination result.

[0009] In a second aspect, an embodiment of the present invention provides a product recommendation model training method, comprising:

[0010] Acquire a sample user feature matrix based on the attribute feature data of the sample users, and acquire a sample product feature matrix based on the attribute feature data of the sample products, wherein the sample products are historically selected products of the sample users;

[0011] Establishing a sample user-product interaction graph based on the selection records of the sample users selecting the sample products, and obtaining a sample adjacency matrix corresponding to the sample user-product interaction graph;

[0012] Inputting the sample user feature matrix, the sample product feature matrix and the sample adjacency matrix into a product recommendation model to be trained, wherein the product recommendation model includes an autoencoder and a graph convolutional neural network;

[0013] By using the autoencoder, attribute features of the sample user and the sample product are correspondingly obtained based on the sample user feature matrix and the sample product feature matrix, thereby obtaining sample user attribute features and sample product attribute features;

[0014] Based on the sample user attribute features, the sample product attribute features and the sample adjacency matrix, determining the sample product recommendation results corresponding to the sample user through the graph convolutional neural network; and

[0015] The function value of the training loss function is determined based on the sample product recommendation results corresponding to the sample users, and the model parameters in the product recommendation model are adjusted according to the function value of the training loss function.

[0016] In a third aspect, an embodiment of the present invention further provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, a product recommendation method or a product recommendation model training method as described in any one of the embodiments of the present invention is implemented.

[0017] In a fourth aspect, an embodiment of the present invention further provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements a product recommendation method or a product recommendation model training method as described in any one of the embodiments of the present invention.

[0018] A product recommendation method provided by an embodiment of the present invention obtains high-quality attribute information features of a target user by utilizing an autoencoder with an advantage in extracting attribute information features, and then transmits the high-quality user attribute information features to a graph convolutional neural network with an advantage in extracting structural information, and determines whether to recommend a candidate product to the target user based on the high-quality attribute information features through the graph convolutional neural network. The method can combine the high-quality attribute information features and the high-quality structural information features extracted by the graph convolutional neural network to fully explore the high-order correlation between the target user and the candidate products, thereby improving the accuracy of recommendations when recommending products to the target user and improving user satisfaction. In addition, since the product recommendation method provided by an embodiment of the present invention does not require interactive matrix splitting, the time complexity is low, and therefore the recommendation efficiency is high.

[0019] The product recommendation model training method further provided in the implementation of the present invention utilizes an autoencoder with an advantage in attribute information feature extraction to obtain high-quality attribute information features of sample users and sample products, and transmits them to a graph convolutional neural network. Thereafter, the graph convolutional neural network is used to obtain sample product recommendation results that combine high-quality attribute information features and high-quality structural information features, and the model parameters of the product recommendation model are adjusted based on the sample product recommendation results. This allows training to obtain a product recommendation model with high recommendation efficiency and high recommendation accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the technical solution of the present invention, the drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0021] Figure 1 is a flowchart of a product recommendation method provided by an embodiment of the present invention;

[0022] Figure 2 is a flowchart of a product recommendation method provided by an embodiment of the present invention;

[0023] Figure 3 is another flowchart of the product recommendation method provided by an embodiment of the present invention;

[0024] Figure 4 is another flowchart of the product recommendation method provided by an embodiment of the present invention;

[0025] Figure 5 is another flowchart of the product recommendation method provided by an embodiment of the present invention;

[0026] Figure 6 is another flowchart of the product recommendation method provided by an embodiment of the present invention;

[0027] Figure 7 is another flowchart of the product recommendation method provided by an embodiment of the present invention;

[0028] Figure 8 It is a flowchart of a product recommendation model training method provided by an embodiment of the present invention;

[0029] Fig. 9 is a structural schematic diagram of a product recommendation device provided by an embodiment of the present invention;

[0030] Fig.10 A structural schematic diagram of a product recommendation model training device provided by an embodiment of the present invention;

[0031] Fig.11 It is a structural schematic diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0032] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0033] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0034] Figure 1 The present invention provides a flowchart of a product recommendation method according to an embodiment of the present invention. The method can be performed by a product recommendation device according to an embodiment of the present invention. The device can be implemented in software and / or hardware. In a specific embodiment, the device can be integrated into an electronic device, such as a computer, a server, etc. The following embodiments will be described by taking the device integrated into an electronic device as an example. Figure 1 , the method may specifically include the following steps:

[0035] Step 101, obtaining a target user feature matrix based on the attribute feature data of the target user. This step can facilitate inputting the target user features into the pre-trained product recommendation model, and further facilitate extracting the attribute information features of the target user feature matrix through the autoencoder of the product recommendation model.

[0036] Specifically, the target users may be users of financial products, and the target users may include old users, new users and / or potential users.

[0037] Specifically, the attribute characteristic data of the target user includes pre-collected user information, which may include information such as number, gender, age, job, education level, annual income and / or deposit amount.

[0038] It should be pointed out that the user information collected is information and data authorized by the user or fully authorized by all parties, and the collection, storage, use, processing, transmission, provision, disclosure and application of relevant data comply with the relevant laws, regulations and standards of relevant countries and regions, take necessary confidentiality measures, do not violate public order and good morals, and provide corresponding operation entrances for users to choose to authorize or refuse.

[0039] Optionally, the process of obtaining the target user feature matrix based on the target user's attribute feature data includes: dimensionlessly processing continuous information such as age, annual income and deposit amount, and further binning or binarizing it to obtain continuous information features; uniquely encoding discrete information representing categories such as education level and gender to obtain discrete information features; and concatenating the above-mentioned continuous information features and discrete information features to obtain the above-mentioned target user feature matrix.

[0040] Step 102: Input the target user feature matrix into a pre-trained product recommendation model, which includes an autoencoder and a graph convolutional neural network. This step can help extract high-quality attribute information features of the target user through the autoencoder with the advantage of extracting attribute information.

[0041] Specifically, the autoencoder may include a multi-layer encoder and a multi-layer decoder, and the parameters of the multi-layer encoder and the multi-layer decoder are the same. The graph convolutional neural network may specifically include a multi-layer graph convolutional network.

[0042] Specifically, the number of layers of the multi-layer encoder and the multi-layer graph convolutional network may be the same or different.

[0043] Step 103, the attribute features of the target user are obtained by using the autoencoder based on the target user feature matrix to obtain the attribute features of the target user. This step can fully utilize the ability of the autoencoder to extract attribute information features, extract high-quality attribute information features of the target user, and facilitate the transfer of the high-quality attribute information features of the target user to the graph convolutional neural network to make up for the lack of the ability of the graph convolutional neural network to extract attribute information features.

[0044] Optionally, the process of obtaining the attribute features of the target user by acquiring the attribute features of the target user based on the target user feature matrix through the autoencoder includes: extracting the attribute feature information of the target user feature matrix through the multi-layer encoder of the autoencoder.

[0045] Specifically, the attribute information of the target user feature matrix may also be extracted by the multi-layer encoder and multi-layer decoder of the above-mentioned autoencoder.

[0046] Specifically, the number of the above target user attribute features may be one or more.

[0047] Optional, such as Figure 2 As shown, the above-mentioned multiple target user attribute features include: through a multi-layer encoder, taking the target user feature matrix as the initial input, and then extracting the attribute features of the target user layer by layer using the output of the previous layer as input, the target user attribute information related features output by each layer of encoder are the output target user features of each encoder.

[0048] Specifically, the encoders at each layer can extract the attribute features of the target user based on the following formula:

[0049]

[0050] Among them, σ(·) represents the activation function, and Relu can be used as the activation function; and They are the trainable weight matrix and bias of the Nth layer encoder respectively; and are the inputs of the N-1 layer and the N layer in the encoder, respectively, and H e (0) is set as the target user feature matrix.

[0051] Specifically, the above-mentioned multiple target user attribute features may also be target user attribute information related features output by a layer in a multi-layer encoder or a multi-layer decoder.

[0052] Specifically, the plurality of target user attribute features include target user attribute information related features output by each layer of decoders when the target user attribute features are extracted layer by layer using the output of the previous layer as input through multi-layer encoders and multi-layer decoders.

[0053] Step 104, based on the attribute features of the target user, determine whether to recommend the candidate product to the target user through the graph convolutional neural network to obtain the recommended product determination result. Step Based on steps 101 to 103, the high-quality attribute information features of the target user are obtained by using an autoencoder with an advantage in extracting attribute information features, and then the high-quality user attribute information features are passed to the graph convolutional neural network with an advantage in extracting structural information, and based on the high-quality attribute information features, the graph convolutional neural network is used to determine whether to recommend the candidate product to the target user. The high-quality attribute information features and the high-quality structural information features extracted by the graph convolutional neural network can be combined to fully explore the high-order correlation between the target user and the candidate product, thereby improving the recommendation accuracy when recommending products to the target user and improving user satisfaction; in addition, since the product recommendation method provided in the embodiment of the present invention does not need to split the interaction matrix, the time complexity is low, and therefore the recommendation efficiency is high.

[0054] It can be understood that the graph convolutional neural network can extract the structural information of the target user's attribute feature data. Based on the target user's attribute features and the structural information of the target user's attribute feature data extracted by the graph convolutional neural network, the characteristics combining the target user's attribute features and structural features can be obtained.

[0055] Specifically, the candidate products may include one product that can be recommended or may include multiple products that can be recommended.

[0056] Optionally, the process of determining whether to recommend a candidate product to a target user based on the target user attribute characteristics by using a graph convolutional neural network to obtain a recommended product determination result includes:

[0057] Based on the adjacency matrix corresponding to the predetermined interaction relationship between the target user and the candidate product and the target user attribute characteristics, the recommended product determination result is determined by the above-mentioned graph convolutional neural network.

[0058] Optionally, when the target user has no interaction record with the candidate product, each element of the adjacency matrix of the interaction relationship between the target user and the candidate product is determined to be 0.

[0059] Optionally, the process of determining whether to recommend a candidate product to a target user based on the target user attribute characteristics by using a graph convolutional neural network to obtain a recommended product determination result includes:

[0060] Based on the attribute characteristics of the target user, the characteristics of the combination of the attribute characteristics and structural characteristics of the target user are determined through a graph convolutional neural network to obtain the comprehensive characteristics of the target user; the recommended product determination result is determined based on the comprehensive characteristics of the target user and the comprehensive characteristics of the candidate products.

[0061] Specifically, the candidate products may all be old products or may include new products, and the comprehensive features of the candidate products may be acquired in advance or may be acquired before step 104 .

[0062] Specifically, when all of the candidate products are old products, the comprehensive characteristics of the candidate products may be acquired in advance; when the candidate products include new products, the comprehensive characteristics of the candidate products may be acquired before step 104 .

[0063] The product recommendation method provided by the embodiment of the present invention is further described below. Figure 3 As shown, the following steps may be included:

[0064] Step 301: Acquire a target user feature matrix based on the attribute feature data of the target user.

[0065] Step 302: Input the target user feature matrix into a pre-trained product recommendation model, where the product recommendation model includes an autoencoder and a graph convolutional neural network.

[0066] Step 303: Obtain the attribute features of the target user based on the target user feature matrix through an autoencoder to obtain the target user attribute features.

[0067] Step 304: establish a target user-product interaction graph based on the target user's selection record of the candidate products, and obtain a target adjacency matrix corresponding to the target user-product interaction graph.

[0068] Optionally, the process of establishing a target user-product interaction graph based on the target user's selection record of the candidate product includes: taking the target user as the source node, taking the candidate product as the target node, and constructing a connection line between the source node and the target node, from the source node to the target node.

[0069] Step 305: Based on the target user attribute characteristics and the target adjacency matrix, determine whether to recommend the candidate product to the target user through a graph convolutional neural network to obtain a recommended product determination result.

[0070] Specifically, step 304 may be executed at any time before step 305 , for example, before step 301 or while step 301 is being executed.

[0071] Optionally, the process of determining whether to recommend a candidate product to a target user based on the target user attribute characteristics and the target adjacency matrix by using a graph convolutional neural network to obtain a recommended product determination result includes:

[0072] The self-circulating adjacency matrix and the degree matrix of the self-circulating adjacency matrix corresponding to the target adjacency matrix are obtained; and based on the attribute characteristics of the target user, the self-circulating adjacency matrix and the degree matrix, a graph convolutional neural network is used to determine whether to recommend candidate products to the target user to obtain a recommended product determination result.

[0073] Optionally, the process of obtaining the self-circulating adjacency matrix corresponding to the target adjacency matrix includes: adding a unit matrix to the target adjacency matrix to obtain the above-mentioned informationized adjacency matrix, which can be specifically performed based on the following formula:

[0074] A=A+I

[0075] Among them, A represents the self-circular adjacency matrix, A represents the target adjacency matrix, and I represents the identity matrix.

[0076] It is understandable that by adding the unit matrix to the adjacency matrix, it is possible to avoid incorporating too much domain information into the process of passing features to the subsequent graph convolutional network layer, which results in too little information being included in the network, thereby better extracting the structural features of each node.

[0077] It can be understood that the embodiments of the present invention can be applied to recommending products to old users by establishing a target adjacency matrix corresponding to the selection records of target users and candidate products, and then mining more high-order associations between users and candidate products through a graph convolutional neural network based on the target adjacency matrix, thereby further improving the accuracy of recommending candidate products to target users.

[0078] The product recommendation method provided by the embodiment of the present invention is further described below. Figure 4 As shown, the following steps may be included:

[0079] Step 401: Acquire a target user feature matrix based on the attribute feature data of the target user.

[0080] Step 402: input the target user feature matrix into a pre-trained product recommendation model, where the product recommendation model includes an autoencoder and a graph convolutional neural network.

[0081] Step 403: Obtain the attribute features of the target user based on the target user feature matrix through an autoencoder to obtain the target user attribute features.

[0082] Step 404, obtaining a candidate product feature matrix based on the attribute feature data of the candidate products, inputting the candidate product feature matrix into the product recommendation model, and obtaining the attribute information features of the candidate products based on the candidate product feature matrix through an autoencoder to obtain the attribute features of the candidate products.

[0083] Specifically, when the candidate product is a financial product, the attribute characteristic data of the candidate product includes a series of characteristics such as the issuer, term, price, return, risk and / or liquidity of the financial product.

[0084] Optional, such as Figure 2 As shown, the above-mentioned multiple candidate product attribute features include: through a multi-layer encoder, taking the candidate product feature matrix as the initial input, and then extracting the attribute features of the candidate products layer by layer using the output of the previous layer as input, the attribute information-related features of the candidate products output by each layer of encoder are the output target user features of each encoder.

[0085] Step 405, based on the target user attribute characteristics, the candidate product attribute characteristics and the target adjacency matrix, determine whether to recommend the candidate product to the target user through a graph convolutional neural network to obtain a recommended product determination result.

[0086] Specifically, step 404 may be performed at any time before step 405 , for example, before step 401 , or while steps 401 to 403 are being performed.

[0087] It is understandable that the embodiments of the present invention are applicable to the scenario of adding new candidate products to the candidate products recommended to users. By using an autoencoder to extract the high-quality attribute features of the candidate products and passing the high-quality attribute features to the graph convolutional neural network, it is possible to obtain a high-order correlation between the new candidate products and the users, thereby improving the accuracy of recommending candidate products to users.

[0088] The product recommendation method provided by the embodiment of the present invention is further described below. Figure 5 As shown, Figure 4 Step 405 in may include the following steps:

[0089] Step 4051, based on the attribute characteristics of the target user and the target adjacency matrix, obtain the characteristics of the combination of the attribute characteristics and the structural characteristics of the target user to obtain the comprehensive characteristics of the target user.

[0090] Step 4052, based on the attribute characteristics of the candidate products and the target adjacency matrix, obtain the characteristics of the combination of the attribute characteristics and the structural characteristics of the candidate products to obtain the comprehensive characteristics of the candidate products.

[0091] Specifically, step 4051 may be executed before step 4052, after step 4052, or while step 4051 is being executed.

[0092] Step 4053, based on the comprehensive characteristics of the target user and the comprehensive characteristics of the candidate products, a recommendation score for recommending the candidate products to the target user is determined, and a recommended product determination result is determined based on the recommendation score.

[0093] Optionally, the process of determining a recommendation score for recommending a candidate product to a target user based on comprehensive characteristics of the target user and comprehensive characteristics of the candidate product includes: using a sigmoid function to calculate a recommendation score for recommending a candidate product to a target user based on comprehensive characteristics of the target user and comprehensive characteristics of the candidate product.

[0094] Optionally, the process of determining the recommended product determination result based on the recommendation score includes: determining whether to recommend the candidate product to the target user based on whether the recommendation score is greater than a preset recommendation score threshold.

[0095] Specifically, when the recommendation score is greater than the recommendation score threshold, the candidate product may be determined as a recommendable product, and when the recommendation score is not greater than the recommendation score threshold, the candidate product may be determined as a non-recommended product.

[0096] Optionally, the process of determining the recommended product determination result based on the recommendation score includes:

[0097] A preset number of candidate products with the largest recommendation scores among multiple candidate products are obtained, and the corresponding candidate products are determined as recommendable products.

[0098] Specifically, after determining the recommendable products, the recommendable candidate products may be directly recommended to the user.

[0099] Optionally, the graph convolutional neural network includes L-layer graph convolutional networks, where L is an integer and 2≤L.

[0100] Optional, such as Figure 2 and Figure 6 As shown, Figure 5 Step 4051 in may include the following steps:

[0101] Step 4051A, using the target user feature matrix as the input target user feature of the first-layer graph convolutional network, and obtaining the output target user feature of the first-layer graph convolutional network through the first-layer graph convolutional network based on the input target user feature and the target adjacency matrix of the first-layer graph convolutional network; and

[0102] Specifically, the input target user features of the above-mentioned graph convolutional network can be understood as the input features of each layer of the graph convolutional network used to extract the comprehensive features of the target user, and the output target user features of the above-mentioned graph convolutional network can be understood as the target user attribute features output by each layer of the graph convolutional network after extraction based on the input target user features.

[0103] Specifically, the input target user features of the first-layer graph convolutional network may also be determined based on the target user feature matrix and the target user attribute features.

[0104] Step 4051B, based on the output target user features, target adjacency matrix and target user attribute features of the first layer graph convolutional network, obtain the comprehensive features of the target user.

[0105] Optional, such as Figure 2 and Figure 6 As shown, Figure 5 Step 4052 in may include the following steps:

[0106] Step 4052A, using the candidate product feature matrix as the input candidate product features of the first-layer graph convolutional network, and based on the input candidate product features of the first-layer graph convolutional network and the target adjacency matrix, obtaining the output candidate product features of the first-layer graph convolutional network through the first-layer graph convolutional network.

[0107] Specifically, the input candidate product features of the above-mentioned graph convolutional network can be understood as the input features of each layer of the graph convolutional network used to extract the comprehensive features of the candidate products, and the output candidate product features of the above-mentioned graph convolutional network can be understood as the candidate product attribute features output by each layer of the graph convolutional network after extraction based on the input candidate product features.

[0108] Specifically, the input candidate product features of the above-mentioned first-layer graph convolutional network can also be determined based on the candidate product feature matrix and the candidate product attribute features.

[0109] Step 4052B, based on the output candidate product features of the first layer graph convolutional network, the target adjacency matrix and the candidate product attribute features, obtain the comprehensive features of the candidate products.

[0110] The embodiment of the present invention provides a specific method for determining the recommended product results through a graph convolutional neural network, which can accurately recommend candidate products to target users.

[0111] The product recommendation method provided by the embodiment of the present invention is further described below.

[0112] Optionally, the autoencoder includes L layers of encoders, the target user attribute features include output target user features of each layer of encoders in the first L-1 layers of encoders; the candidate product attribute features include output candidate product features of each layer of encoders in the first L-1 layers of encoders.

[0113] Optional, such as Figure 7 and Figure 2 As shown, Figure 6 Step 4051B in may include:

[0114] Step 4051B1, based on the output target user features of the N-1 layer encoder, the output target user features of the N-1 layer graph convolutional network and the target adjacency matrix, determine the input target user features of the N-1 layer graph convolutional network, where N is an integer and 2≤N≤L.

[0115] Step 4051B2, based on the input target user features of the N-th layer graph convolutional network, the output target user features of the N-th layer graph convolutional network are obtained through the N-th layer graph convolutional network.

[0116] Step 4051B3, obtain the output target user features of the L-th layer graph convolutional network to obtain the target user comprehensive features.

[0117] Optional, such as Figure 7 and Figure 2 As shown, Figure 6 Step 4052B in may include:

[0118] Step 4052B1, based on the output candidate product features of the N-1 layer encoder, the output candidate product features of the N-1 layer graph convolutional network and the target adjacency matrix, determine the input candidate product features of the N-1 layer graph convolutional network.

[0119] Step 4052B2, based on the input candidate product features of the N-th layer graph convolutional network, the output candidate product features of the N-th layer graph convolutional network are obtained through the N-th layer graph convolutional network.

[0120] Step 4052B3, obtain the output candidate product features of the L-th layer graph convolutional network to obtain the comprehensive features of the candidate products.

[0121] Optionally, the process of determining the input target user features of the N-th layer graph convolutional network based on the output target user features of the N-1-th layer encoder, the output target user features of the N-1-th layer graph convolutional network, and the target adjacency matrix includes: determining the input target user features of the N-th layer graph convolutional network based on the output target user features of the N-1-th layer encoder, the output target user features of the N-1-th layer graph convolutional network, the self-circular adjacency matrix and the degree matrix.

[0122] Optionally, the process of determining the input candidate product features of the N-th layer graph convolutional network based on the output candidate product features of the N-1-th layer encoder, the output candidate product features of the N-1-th layer graph convolutional network and the target adjacency matrix includes: determining the input candidate product features of the N-th layer graph convolutional network based on the output candidate product features of the N-1-th layer encoder, the output candidate product features of the N-1-th layer graph convolutional network, the self-circular adjacency matrix and the degree matrix.

[0123] Optionally, the input target user features of the N-th layer graph convolutional network, or the input candidate product features of the N-th layer graph convolutional network can be expressed as

[0124]

[0125] in, Z (N-1)represents the output target user features or output candidate product features of the N-th layer graph convolutional network; A represents the self-circular adjacency matrix; D represents the degree matrix, that is, W (N-1) Represents the weight matrix of the N-th layer graph convolutional network; represents the output user product features or output candidate product features of the N-1th layer encoder; p represents a transfer parameter, which is used to balance the ratio between the output user product features or output candidate product features of the N-1th layer encoder and the output user product features or output candidate product features of the N-1th layer graph convolutional network.

[0126] Specifically, the transfer parameter p can be determined based on empirical data or through experiments, and can be specifically set to 0.5.

[0127] Optionally, the process of obtaining the output target user features of the N-th graph convolutional network through the N-th graph convolutional network based on the input target user features of the N-th graph convolutional network, or obtaining the output candidate product features of the N-th graph convolutional network through the N-th graph convolutional network based on the input candidate product features of the N-th graph convolutional network, can be performed based on the following formula:

[0128]

[0129] Among them, σ(·) represents the activation function.

[0130] Optionally, the aforementioned process of obtaining the output target user features of the first-layer graph convolutional network through the first-layer graph convolutional network based on the input target user features and the target adjacency matrix of the first-layer graph convolutional network, or the process of obtaining the output candidate product features of the first-layer graph convolutional network through the first-layer graph convolutional network based on the input candidate product features and the target adjacency matrix of the first-layer graph convolutional network, can be performed based on the following formula:

[0131]

[0132] Optionally, the aforementioned process of determining the recommendation score for recommending the candidate product to the target user based on the comprehensive characteristics of the target user and the comprehensive characteristics of the candidate product may be performed based on the following formula:

[0133]

[0134] Among them, σ(·) represents the sigmoid function, represents the comprehensive characteristics of target user i, Represents the comprehensive characteristics of candidate product j.

[0135] Specifically, the present invention implements the transmission of the attribute features of the target user or candidate product output by the N-1th layer encoder to the Nth layer graph convolutional network, and can transmit different attribute information output by different encoder layers to the graph convolutional network, thereby increasing the diversity of the attribute information, and further increasing the richness of the corresponding attribute information in the comprehensive features of the target user or the comprehensive features of the candidate product, so as to improve the accuracy of recommending candidate products to users.

[0136] Figure 8 The present invention provides a flowchart of a product recommendation model training method according to an embodiment of the present invention. The method can be performed by a product recommendation model training device according to an embodiment of the present invention. The device can be implemented in software and / or hardware. In a specific embodiment, the device can be integrated into an electronic device, such as a computer, a server, etc. The following embodiments will be described by taking the device integrated into an electronic device as an example. Figure 8 , the method may specifically include the following steps:

[0137] Step 801 , obtaining a sample user feature matrix based on the attribute feature data of the sample users, and obtaining a sample product feature matrix based on the attribute feature data of the sample products, where the sample products are historically selected products of the sample users.

[0138] Step 802: Establish a sample user-product interaction graph based on the selection records of sample users selecting sample products, and obtain a sample adjacency matrix corresponding to the sample user-product interaction graph.

[0139] Step 803: input the sample user feature matrix, the sample product feature matrix and the sample adjacency matrix into the product recommendation model to be trained, where the product recommendation model includes an autoencoder and a graph convolutional neural network.

[0140] Step 804, using an autoencoder to obtain attribute features of sample users and sample products based on a sample user feature matrix and a sample product feature matrix, respectively, to obtain sample user attribute features and sample product attribute features.

[0141] Step 805: Based on the sample user attribute features, the sample product attribute features and the sample adjacency matrix, the sample product recommendation results corresponding to the sample user are determined through a graph convolutional neural network.

[0142] Optionally, the graph convolutional neural network includes L-layer graph convolutional networks, where L is an integer and 2≤L.

[0143] Optionally, the autoencoder includes L layers of encoders, the sample user attribute features include output sample user features of each layer of encoders in the first L-1 layers of encoders; the sample product attribute features include output sample product features of each layer of encoders in the first L-1 layers of encoders.

[0144] Optionally, the process of determining the sample product recommendation result corresponding to the sample user through the graph convolutional neural network based on the sample user attribute features, the sample product attribute features and the sample adjacency matrix includes:

[0145] Based on the attribute characteristics of sample users and the target adjacency matrix, the characteristics of the combination of the attribute characteristics and the structural characteristics of the sample users are obtained to obtain the comprehensive characteristics of the sample users; based on the attribute characteristics of sample products and the target adjacency matrix, the characteristics of the combination of the attribute characteristics and the structural characteristics of the sample products are obtained to obtain the comprehensive characteristics of the sample products; and based on the comprehensive characteristics of sample users and the comprehensive characteristics of sample products, the recommendation scores of the sample products recommended by the sample users are determined, and the sample product recommendation results corresponding to the sample users are determined from the candidate sample products based on the recommendation scores.

[0146] Optionally, the process of obtaining the combined features of the attribute features and the structural features of the sample users based on the attribute features of the sample users and the target adjacency matrix to obtain the comprehensive features of the sample users includes:

[0147] The sample user feature matrix is ​​used as the input sample user features of the first-layer graph convolutional network, and based on the input sample user features and the target adjacency matrix of the first-layer graph convolutional network, the output sample user features of the first-layer graph convolutional network are obtained through the first-layer graph convolutional network; and based on the output sample user features of the first-layer graph convolutional network, the target adjacency matrix and the sample user attribute features, the comprehensive features of the sample users are obtained.

[0148] Optionally, the process of obtaining comprehensive features of sample users based on the output sample user features, target adjacency matrix, and sample user attribute features of the first-layer graph convolutional network includes:

[0149] Based on the output sample user features of the N-1th layer encoder, the output sample user features of the N-1th layer graph convolutional network and the target adjacency matrix, the input sample user features of the Nth layer graph convolutional network are determined, where N is an integer and 2≤N≤L; based on the input sample user features of the Nth layer graph convolutional network, the output sample user features of the Nth layer graph convolutional network are obtained through the Nth layer graph convolutional network; and the output sample user features of the Lth layer graph convolutional network are obtained to obtain the comprehensive features of the sample users.

[0150] Optionally, the process of obtaining the combined features of the attribute features and the structural features of the sample product based on the attribute features of the sample product and the target adjacency matrix to obtain the comprehensive features of the sample product includes:

[0151] Using the sample product feature matrix as the input sample product features of the first-layer graph convolutional network, and obtaining the output sample product features of the first-layer graph convolutional network through the first-layer graph convolutional network based on the input sample product features of the first-layer graph convolutional network and the target adjacency matrix; and

[0152] Based on the output sample product features, target adjacency matrix and sample product attribute features of the first-layer graph convolutional network, the comprehensive features of the sample products are obtained.

[0153] Optionally, the process of obtaining comprehensive features of the sample products based on the output sample product features, the target adjacency matrix, and the sample product attribute features of the first-layer graph convolutional network includes:

[0154] Determine the input sample product features of the N-th layer graph convolutional network based on the output sample product features of the N-1th layer encoder, the output sample product features of the N-1th layer graph convolutional network, and the target adjacency matrix, where N is an integer and 2≤N≤L;

[0155] Obtaining output sample product features of the N-th layer graph convolutional network through the N-th layer graph convolutional network based on the input sample product features of the N-th layer graph convolutional network; and

[0156] The output sample product features of the L-th layer graph convolutional network are obtained to obtain the comprehensive features of the sample products.

[0157] Step 806: determine the function value of the training loss function based on the sample product recommendation results corresponding to the sample users, and adjust the model parameters in the product recommendation model according to the function value of the training loss function.

[0158] Optionally, the above training loss function is a combination of the reconstruction loss of the autoencoder and the cross entropy loss of the graph convolutional neural network, which can be specifically expressed as:

[0159] L=αL sup +βL AE

[0160] Among them, α and β represent weighting coefficients, L sup represents the cross entropy loss of graph convolutional neural network, L AE represents the reconstruction loss of the autoencoder.

[0161] It can be understood that by optimizing the model parameters of the product recommendation model based on the loss function obtained by combining the reconstruction loss of the autoencoder and the cross entropy loss of the graph convolutional neural network, the best product recommendation model can be obtained.

[0162] The product recommendation model training method further provided in the implementation of the present invention utilizes an autoencoder with an advantage in attribute information feature extraction to obtain high-quality attribute information features of sample users and sample products, and transmits them to a graph convolutional neural network. Thereafter, the graph convolutional neural network is used to obtain sample product recommendation results that combine high-quality attribute information features and high-quality structural information features, and the model parameters of the product recommendation model are adjusted based on the sample product recommendation results. This allows training to obtain a product recommendation model with high recommendation efficiency and high recommendation accuracy.

[0163] Fig. 9 is a structural diagram of a product recommendation device provided by an embodiment of the present invention, and the device is suitable for executing the product recommendation method provided by an embodiment of the present invention. Fig. 9 As shown, the device may specifically include:

[0164] The recommendation feature matrix acquisition module 901 is used to acquire a target user feature matrix based on the attribute feature data of the target user.

[0165] The recommendation input module 902 is used to input the target user feature matrix into a pre-trained product recommendation model, which includes an autoencoder and a graph convolutional neural network.

[0166] The recommended attribute feature acquisition module 903 is used to obtain the attribute features of the target user based on the target user feature matrix through an autoencoder to obtain the attribute features of the target user.

[0167] The recommendation result acquisition module 904 is used to determine whether to recommend a candidate product to the target user based on the attribute characteristics of the target user through a graph convolutional neural network to obtain a recommended product determination result.

[0168] The product recommendation device provided by the embodiment of the present invention can obtain high-quality attribute information features of the target user by utilizing an autoencoder with an advantage in extracting attribute information features, and then pass the high-quality user attribute information features to a graph convolutional neural network with an advantage in extracting structural information, and determine whether to recommend candidate products to the target user based on the high-quality attribute information features through the graph convolutional neural network. It can combine the high-quality attribute information features and the high-quality structural information features extracted by the graph convolutional neural network to fully explore the high-order correlation between the target user and the candidate products, thereby improving the recommendation accuracy when recommending products to the target user and improving user satisfaction. In addition, since the product recommendation method provided by the embodiment of the present invention does not require interactive matrix splitting and has low time complexity, the recommendation efficiency is relatively high.

[0169] Fig.10 is a structural diagram of a product recommendation model training device provided by an embodiment of the present invention, and the device is suitable for executing the product recommendation model training method provided by an embodiment of the present invention. Fig.10 As shown, the device may specifically include:

[0170] The sample feature matrix acquisition module 1001 is used to acquire a sample user feature matrix based on the attribute feature data of the sample user, and to acquire a sample product feature matrix based on the attribute feature data of the sample product, where the sample product is a product selected historically by the sample user.

[0171] The sample adjacency matrix acquisition module 1002 is used to establish a sample user-product interaction graph based on the selection records of sample users selecting sample products, and to acquire a sample adjacency matrix corresponding to the sample user-product interaction graph.

[0172] The sample input module 1003 is used to input the sample user feature matrix, the sample product feature matrix and the sample adjacency matrix into the product recommendation model to be trained, and the product recommendation model includes an autoencoder and a graph convolutional neural network.

[0173] The sample attribute feature extraction module 1004 is used to obtain the attribute features of the sample users and the sample products based on the sample user feature matrix and the sample product feature matrix respectively through an autoencoder to obtain the sample user attribute features and the sample product attribute features.

[0174] The sample product recommendation result acquisition module 1005 is used to determine the sample product recommendation result corresponding to the sample user through a graph convolutional neural network based on the sample user attribute characteristics, the sample product attribute characteristics and the sample adjacency matrix.

[0175] The parameter adjustment module 1006 is used to determine the function value of the training loss function based on the sample product recommendation results corresponding to the sample users, and adjust the model parameters in the product recommendation model according to the function value of the training loss function.

[0176] The product recommendation model training device provided in the embodiment of the present invention can utilize the autoencoder with the advantage of attribute information feature extraction to obtain high-quality attribute information features of sample users and sample products, and pass them to the graph convolutional neural network. Thereafter, the graph convolutional neural network is used to obtain the sample product recommendation results that combine the high-quality attribute information features and the high-quality structural information features, and the model parameters of the product recommendation model are adjusted based on the sample product recommendation results. A product recommendation model with high recommendation efficiency and high recommendation accuracy can be trained.

[0177] Those skilled in the art can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional modules is used as an example for illustration. In actual applications, the above-mentioned functions can be assigned to different functional modules as needed, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above. The specific working process of the functional modules described above can refer to the corresponding process in the aforementioned method embodiment.

[0178] In addition, the product recommendation device provided by the embodiment of the present invention can execute the product recommendation method provided by any embodiment of the present invention, and the product recommendation model training device provided by the embodiment of the present invention can execute the product recommendation model training method provided by any embodiment of the present invention, which will not be repeated here.

[0179] An embodiment of the present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the product recommendation method or product recommendation model training method provided in any of the above embodiments is implemented.

[0180] An embodiment of the present invention further provides a computer-readable medium having a computer program stored thereon, and when the program is executed by a processor, the product recommendation method or product recommendation model training method provided in any of the above embodiments is implemented.

[0181] The embodiment of the present invention further provides a computer program product, including a computer program, which, when executed by a processor, implements the product recommendation method or product recommendation model training method as described in any one of the embodiments of the present invention.

[0182] Reference below Fig.11 , which shows a schematic diagram of the structure of a computer system 1100 of an electronic device suitable for implementing an embodiment of the present invention. Fig.11 The electronic device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention.

[0183] like Fig.11 As shown, the computer system 1100 includes a central processing unit (CPU) 1101, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 1102 or a program loaded from a storage part 1108 into a random access memory (RAM) 1103. In the RAM 1103, various programs and data required for the operation of the system 1100 are also stored. The CPU 1101, the ROM 1102, and the RAM 1103 are connected to each other via a bus 1104. An input / output (I / O) interface 1105 is also connected to the bus 1104.

[0184] The following components are connected to the I / O interface 1105: an input section 1106 including a keyboard, a mouse, etc.; an output section 1107 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 1108 including a hard disk, etc.; and a communication section 1109 including a network interface card such as a LAN card, a modem, etc. The communication section 1109 performs communication processing via a network such as the Internet. A drive 1110 is also connected to the I / O interface 1105 as needed. A removable medium 1111, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 1110 as needed, so that a computer program read therefrom is installed into the storage section 1108 as needed.

[0185] In particular, according to the embodiments disclosed in the present invention, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the embodiments disclosed in the present invention include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program contains program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network through the communication part 1109, and / or installed from the removable medium 1111. When the computer program is executed by the central processing unit (CPU) 1101, the above-mentioned functions defined in the system of the present invention are executed.

[0186] It should be noted that the computer-readable medium shown in the present invention may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present invention, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, device or device. In the present invention, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries a computer-readable program code. This propagated data signal may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer-readable signal media may also be any computer-readable medium other than computer-readable storage media, which may send, propagate or transmit a program for use by or in conjunction with an instruction execution system, apparatus or device. The program code contained on the computer-readable medium may be transmitted using any appropriate medium, including but not limited to: wireless, wire, optical cable, RF, etc., or any suitable combination of the above.

[0187] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present invention. In this regard, each box in the flow chart or block diagram can represent a module, a program segment, or a part of a code, and the above-mentioned module, program segment, or a part of a code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flow chart, and the combination of the boxes in the block diagram or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0188] The modules and / or units involved in the embodiments of the present invention may be implemented in software or hardware. The modules and / or units described may also be provided in a processor, for example, may be described as: a processor including a recommendation feature matrix acquisition module, a recommendation input module, a recommendation attribute feature acquisition module, and a recommendation result acquisition module; or, may be described as: a processor including a sample feature matrix acquisition module, a sample adjacency matrix acquisition module, a sample input module, and a sample attribute feature extraction module. The names of these modules do not, in some cases, constitute limitations on the modules themselves.

[0189] As another aspect, the present invention also provides a computer-readable medium, which may be included in the device described in the above embodiment; or it may exist independently without being assembled into the device. The above computer-readable medium carries one or more programs, and when the above one or more programs are executed by a device, the device implements: obtaining a target user feature matrix based on the attribute feature data of the target user; inputting the target user feature matrix into a pre-trained product recommendation model, the product recommendation model includes an autoencoder and a graph convolutional neural network; obtaining the attribute features of the target user based on the target user feature matrix through the autoencoder to obtain the target user attribute features; and determining whether to recommend a candidate product to the target user based on the target user attribute features through the graph convolutional neural network to obtain a recommended product determination result.

[0190] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions may occur depending on design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A product recommendation method, characterized in that: include: Acquire a target user feature matrix based on the attribute feature data of the target user; Inputting the target user feature matrix into a pre-trained product recommendation model, wherein the product recommendation model includes an autoencoder and a graph convolutional neural network; Obtaining the attribute features of the target user based on the target user feature matrix by the autoencoder to obtain the target user attribute features; and Based on the attribute characteristics of the target user, the graph convolutional neural network is used to determine whether to recommend a candidate product to the target user to obtain a recommended product determination result.

2. The product recommendation method according to claim 1, characterized in that: Before determining whether to recommend a candidate product to the target user based on the target user attribute characteristics by the graph convolutional neural network to obtain a recommended product determination result, the product recommendation method further includes: Establishing a target user-product interaction graph based on the selection record of the target user selecting the candidate product, and obtaining a target adjacency matrix corresponding to the target user-product interaction graph; The determining whether to recommend a candidate product to the target user based on the target user attribute characteristics by using the graph convolutional neural network to obtain a recommended product determination result includes: Based on the target user attribute characteristics and the target adjacency matrix, the graph convolutional neural network is used to determine whether to recommend a candidate product to the target user to obtain a recommended product determination result.

3. The product recommendation method according to claim 2, characterized in that: Before determining whether to recommend a candidate product to the target user based on the target user attribute characteristics and the target adjacency matrix by the graph convolutional neural network to obtain a recommended product determination result, the product recommendation method further includes: Obtaining a candidate product feature matrix based on the attribute feature data of the candidate products, inputting the candidate product feature matrix into the product recommendation model, and obtaining the attribute information features of the candidate products based on the candidate product feature matrix through the autoencoder to obtain the attribute features of the candidate products; The determining whether to recommend a candidate product to the target user based on the target user attribute characteristics and the target adjacency matrix by using the graph convolutional neural network to obtain a recommended product determination result includes: Based on the target user attribute features, the candidate product attribute features and the target adjacency matrix, the graph convolutional neural network is used to determine whether to recommend the candidate product to the target user to obtain a recommended product determination result.

4. The product recommendation method according to claim 3, characterized in that: The determining whether to recommend a candidate product to the target user based on the target user attribute features, the candidate product attribute features and the target adjacency matrix by the graph convolutional neural network to obtain a recommended product determination result includes: Based on the target user attribute characteristics and the target adjacency matrix, a characteristic combining the attribute characteristics and the structural characteristics of the target user is obtained to obtain a comprehensive characteristic of the target user; Based on the attribute characteristics of the candidate products and the target adjacency matrix, obtaining a combination of the attribute characteristics and the structural characteristics of the candidate products to obtain comprehensive characteristics of the candidate products; and A recommendation score for recommending the candidate product to the target user is determined based on the comprehensive characteristics of the target user and the comprehensive characteristics of the candidate product, and the recommended product determination result is determined based on the recommendation score.

5. The product recommendation method according to claim 4, characterized in that: The graph convolutional neural network includes an L-layer graph convolutional network, where L is an integer and 2≤L; The step of obtaining a combination of the attribute characteristics and the structural characteristics of the target user based on the attribute characteristics of the target user and the target adjacency matrix to obtain the comprehensive characteristics of the target user includes: Using the target user feature matrix as an input target user feature of the first-layer graph convolutional network, and obtaining an output target user feature of the first-layer graph convolutional network through the first-layer graph convolutional network based on the input target user feature of the first-layer graph convolutional network and the target adjacency matrix; and Based on the output target user features of the first layer graph convolutional network, the target adjacency matrix and the target user attribute features, obtaining the target user comprehensive features; The step of obtaining a combination of the attribute characteristics and the structural characteristics of the candidate product based on the attribute characteristics of the candidate product and the target adjacency matrix to obtain the comprehensive characteristics of the candidate product includes: Using the candidate product feature matrix as input candidate product features of the first-layer graph convolutional network, and obtaining output candidate product features of the first-layer graph convolutional network through the first-layer graph convolutional network based on the input candidate product features of the first-layer graph convolutional network and the target adjacency matrix; and Based on the output candidate product features of the first layer graph convolutional network, the target adjacency matrix and the candidate product attribute features, the comprehensive features of the candidate product are obtained.

6. The product recommendation method according to claim 5, characterized in that: The autoencoder includes L layers of encoders, the target user attribute features include output target user features of each layer of encoders in the first L-1 layers of encoders; the candidate product attribute features include output candidate product features of each layer of encoders in the first L-1 layers of encoders; The output target user features, the target adjacency matrix and the target user attribute features based on the first layer of graph convolutional network are used to obtain the target user comprehensive features, including: Determine the input target user features of the Nth layer graph convolutional network based on the output target user features of the N-1th layer encoder, the output target user features of the N-1th layer graph convolutional network, and the target adjacency matrix, where N is an integer and 2≤N≤L; Based on the input target user features of the N-th layer graph convolutional network, the output target user features of the N-th layer graph convolutional network are obtained through the N-th layer graph convolutional network; Obtaining the output target user features of the L-th layer graph convolutional network to obtain the target user comprehensive features; The output candidate product features based on the first layer graph convolutional network, the target adjacency matrix and the candidate product attribute features, obtaining the candidate product comprehensive features, includes: Determine the input candidate product features of the Nth layer graph convolutional network based on the output candidate product features of the N-1th layer encoder, the output candidate product features of the N-1th layer graph convolutional network, and the target adjacency matrix, where N is an integer and 2≤N≤L; Based on the input candidate product features of the N-th layer graph convolutional network, the output candidate product features of the N-th layer graph convolutional network are obtained through the N-th layer graph convolutional network; The output candidate product features of the L-th layer graph convolutional network are obtained to obtain the comprehensive features of the candidate products.

7. The product recommendation method according to claim 2, characterized in that: The determining whether to recommend a candidate product to the target user based on the target user attribute characteristics and the target adjacency matrix by using the graph convolutional neural network to obtain a recommended product determination result includes: Obtaining a self-loop adjacency matrix corresponding to the target adjacency matrix and a degree matrix of the self-loop adjacency matrix; and Based on the target user attribute characteristics, the self-circulating adjacency matrix and the degree matrix, the graph convolutional neural network is used to determine whether to recommend a candidate product to the target user to obtain a recommended product determination result.

8. A product recommendation model training method, characterized in that: include: Acquire a sample user feature matrix based on the attribute feature data of the sample users, and acquire a sample product feature matrix based on the attribute feature data of the sample products, wherein the sample products are historically selected products of the sample users; Establishing a sample user-product interaction graph based on the selection records of the sample users selecting the sample products, and obtaining a sample adjacency matrix corresponding to the sample user-product interaction graph; Inputting the sample user feature matrix, the sample product feature matrix and the sample adjacency matrix into a product recommendation model to be trained, wherein the product recommendation model includes an autoencoder and a graph convolutional neural network; By using the autoencoder, attribute features of the sample user and the sample product are correspondingly obtained based on the sample user feature matrix and the sample product feature matrix, thereby obtaining sample user attribute features and sample product attribute features; Based on the sample user attribute features, the sample product attribute features and the sample adjacency matrix, determining the sample product recommendation results corresponding to the sample user through the graph convolutional neural network; as well as The function value of the training loss function is determined based on the sample product recommendation results corresponding to the sample users, and the model parameters in the product recommendation model are adjusted according to the function value of the training loss function.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the product recommendation method according to any one of claims 1 to 7 or the product recommendation model training method according to claim 8 is implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, it implements the product recommendation method as described in any one of claims 1 to 7 or the product recommendation model training method as described in claim 8.

Images