Social recommendation method and device based on heterogeneous hypergraph and meta-path

Abstract

The invention belongs to the technical field of nuclear power, and particularly relates to a social recommendation method and device based on a heterogeneous hypergraph and a meta-path. According to the social recommendation method based on the heterogeneous hypergraph and the meta-path, implicit social relations are mined to enrich an original social network, the performance of a social recommendation system is trying to be further improved, but the implicit social relations are often not reliable, but can bring much noise, and the social recommendation effect is poor. The problem can be solved to a certain extent by introducing the credibility of the user; utilizing the hypergraph to learn complex high-order relationships in the graph, wherein the complex high-order relationships comprise a user-article interaction relationship and a social relationship; the two features are respectively user in-degree features and user out-degree features, the user in-degree features describe the influence of implicit feedback, and the user out-degree features describe the influence of trust propagation. Therefore, the recommendation performance of the social recommendation system is further improved by mining the implicit social relationship, constructing the hypergraph learning high-order relationship and describing the fine-grained user feature representation.

Description

Technical Field

[0001] The present invention belongs to the field of nuclear power technology, and specifically relates to a social recommendation method and device based on heterogeneous hypergraph and meta-path. Background Art

[0002] Recommendation systems mine user interests from large amounts of data by analyzing the associations between users and items. Traditional recommendation systems use collaborative filtering and historical interaction data between users and items to model users and items, and finally obtain unknown predicted scores. Influenced by some social communication theories, users with social connections are more likely to show similar interests and hobbies. For example, colleagues often share on the Taobao platform. Therefore, using social relationships to improve the performance of recommendation systems has become increasingly popular in academia and industry.

[0003] However, existing social recommendation systems still face three major challenges. First, when faced with sparse and unbalanced social networks, the performance of social recommendation systems will still be limited; second, most existing social recommendation systems only consider low-order neighbor information aggregation, but high-order beneficial information is usually ignored. For example, two users who have no direct relationship in real life may show similar interests and hobbies due to the existence of potential social relationship chains, but such complex social relationships are usually difficult to use directly. Therefore, how to obtain high-order beneficial social information has become a major difficulty; third, almost all social recommendation systems ignore the potential feedback effect of the trustor on the trustee. The trustee may change his or her opinion on a certain item based on the trustor's feedback. Therefore, it is urgent to improve the accuracy of the social recommendation system to make the recommended content more targeted. Summary of the invention

[0004] In order to overcome the problems existing in the related art, a social recommendation method and device based on heterogeneous hypergraph and meta-path are provided.

[0005] According to one aspect of an embodiment of the present disclosure, a social recommendation method based on heterogeneous hypergraph and meta-path is provided, the method comprising:

[0006] Step 1: Mining implicit relationships F between users U Implicit relationship F between items I ;

[0007] Step 2, determining the user trust matrix B according to the 0-1 trust relationship between users and the topological structure matrix T;

[0008] Step 3: Divide user communities. Different user communities have different interest distributions. A hypergraph is constructed based on the divided user communities. A user-item interaction graph G is constructed based on the user historical interaction data.R =<U∪V,A> , where U and V represent the user set and item set respectively, A is transformed from the user-item interaction matrix R, where M represents the number of users and N represents the number of items:

[0009]

[0010] Step 4: Construct the user social network graph G based on the user trust level B S =<U,B> ;

[0011] Step 5: Transform the user-item interaction graph G R With the user social network graph G S The fusion is a fusion graph F;

[0012] Step 6: Randomly select the most influential K users on the trust matrix B;

[0013] Step 7: Randomly walk on the fusion graph F to obtain the correlation coefficient matrix S between the most influential K users and their neighbor nodes = [s i,k ] (M+N)×K ;

[0014] Step 8, divide the user community according to the correlation coefficient matrix S and construct the hypergraph matrix H;

[0015] Step 9: Pre-train the hypergraph matrix H to learn the high-order information of the nodes. After pre-training, the initial embedding representation P of the user and the item is obtained. M×d and Q N×d , d represents the embedding dimension;

[0016] Step 10: F U and F I Integrate into the fusion graph F to generate a graph attention network G, in which meta-paths based on user nodes and item nodes are defined. For user nodes, the user in-degree features are defined based on the in-degree features and out-degree features. The user in-degree features are divided into 5 types of meta-paths, the user out-degree features are also divided into 5 types of meta-paths, and the meta-paths of item nodes are divided into 3 types.

[0017] Step 11: Send the five types of meta-paths of the user's in-degree features into the meta-path-based graph attention network to obtain five embeddings with different meanings ( and );

[0018] Step 12: Use a multi-layer perceptron to fuse these five embeddings with different meanings to obtain the final in-degree feature representation of user i.

[0019] Step 13: Get the out-degree feature representation of user i

[0020] Step 14: Use a multi-layer perceptron to fuse the in-degree feature representation of user i And out-degree feature representation Get the final feature representation p of user i i ;

[0021] Step 15: Send the three types of meta-paths of item j’s features into the meta-path-based graph attention network to obtain three embeddings (q j1 ,q j2 and q j3 );

[0022] Step 16: Use a multi-layer perceptron to fuse these three embeddings with different meanings to obtain the final feature representation q of item j. j ;

[0023] Step 17: Predict user i’s rating of item j through a multi-layer perceptron

[0024]

[0025] Step 18: Based on multiple Create a recommendation list.

[0026] In one possible implementation, a linear model is used to predict stable ratings The final score of user i on item j is expressed as And generate a recommendation list.

[0027] In a possible implementation, a Centerloss loss term is used to constrain the same user community.

[0028] In one possible implementation, in the graph attention network, a dropnode mechanism is used to filter node meta-paths.

[0029] According to one aspect of an embodiment of the present disclosure, a social recommendation device based on heterogeneous hypergraphs and meta-paths is provided, the device comprising:

[0030] Mining module, used to mine implicit relationships between users F U Implicit relationship F between items I ;

[0031] A first determination module is used to determine a user trust degree matrix B according to a 0-1 trust relationship between users and a topological structure matrix T;

[0032] The partitioning module is used to divide user communities. Different user communities have different interest distributions. A hypergraph is constructed based on the divided user communities. A user-item interaction graph G is constructed based on the user's historical interaction data. R =<U∪V,A> , where U and V represent the user set and item set respectively, A is transformed from the user-item interaction matrix R, where M represents the number of users and N represents the number of items:

[0033]

[0034] The first building module is used to build a user social network graph G based on the user trust level B. S =<U,B> ;

[0035] The first fusion module is used to transform the user-item interaction graph G R With the user social network graph G S The fusion is a fusion graph F;

[0036] The first selection module is used to randomly select the most influential K users on the trust matrix B;

[0037] The second selection module is used to randomly walk on the fusion graph F to obtain the correlation coefficient matrix S = [s i,k ] (M+N)×K ;

[0038] The second construction module is used to divide the user community according to the correlation coefficient matrix S and construct the hypergraph matrix H;

[0039] The training module is used to learn the high-order information of nodes through pre-training according to the hypergraph matrix H. After pre-training, the initial embedding representation P of users and items is obtained. M×d and Q N×d , d represents the embedding dimension;

[0040] Generate a module for F U and F I Integrate into the fusion graph F to generate a graph attention network G, in which meta-paths based on user nodes and item nodes are defined. For user nodes, the user in-degree features are defined based on the in-degree features and out-degree features. The user in-degree features are divided into 5 types of meta-paths, the user out-degree features are also divided into 5 types of meta-paths, and the meta-paths of item nodes are divided into 3 types.

[0041] The second determination module is used to send the five types of meta-paths of the user's in-degree features into the meta-path-based graph attention network to obtain five embeddings with different meanings ( and );

[0042] The user in-degree feature module is used to fuse these five embeddings with different meanings through a multi-layer perceptron to obtain the final in-degree feature representation of user i.

[0043] User out-degree feature module, used to obtain the out-degree feature representation of user i

[0044] User feature representation module, used to fuse the in-degree feature representation of user i through a multi-layer perceptron And out-degree feature representation Get the final feature representation p of user i i ;

[0045] The item feature determination module is used to send the three types of meta-paths of item j features into the meta-path-based graph attention network to obtain three embeddings (q j1 ,q j2 and q j3 );

[0046] The item feature representation module is used to fuse these three embeddings with different meanings through a multi-layer perceptron to obtain the final feature representation q of item j. j ;

[0047] The rating module is used to predict the rating of user i for item j through a multi-layer perceptron.

[0048]

[0049] The recommendation list determination module is used to determine the recommended list based on multiple Create a recommendation list.

[0050] In one possible implementation, a linear model is used to predict stable ratings The final score of user i on item j is expressed as And generate a recommendation list.

[0051] In a possible implementation, a Centerloss loss term is used to constrain the same user community.

[0052] In one possible implementation, in the graph attention network, a dropnode mechanism is used to filter node meta-paths.

[0053] According to another aspect of an embodiment of the present disclosure, a social recommendation device based on heterogeneous hypergraph and meta-path is provided, the device comprising:

[0054] processor;

[0055] a memory for storing processor-executable instructions;

[0056] Wherein, the processor is configured to execute the above method.

[0057] According to another aspect of an embodiment of the present disclosure, a non-volatile computer-readable storage medium is provided, on which computer program instructions are stored, and the computer program instructions implement the above method when executed by a processor.

[0058] The beneficial effects of the present disclosure are: the social recommendation method based on heterogeneous hypergraph and meta-path provided by the present disclosure mines implicit social relations to enrich the original social network, and attempts to further improve the performance of the social recommendation system. However, these implicit social relations are often not so reliable and will bring a lot of noise. Introducing the credibility of users can solve this problem to a certain extent; using hypergraphs to learn complex high-order relations in the graph, including user-item interaction relations and social relations; starting from user features from two perspectives, namely user in-degree features and user out-degree features, the user in-degree features characterize the influence of implicit feedback, and the user out-degree features characterize the influence of trust propagation. Thus, by mining implicit social relations, constructing hypergraphs to learn high-order relations, and characterizing fine-grained user feature representations, the recommendation performance of the social recommendation system is further improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 It is a flowchart of a social recommendation method based on heterogeneous hypergraph and meta-path shown in an embodiment of the present disclosure.

[0060] Figure 2 It is a flow chart of a social recommendation device based on heterogeneous hypergraph and meta-path shown in an embodiment of the present disclosure. DETAILED DESCRIPTION

[0061] The present disclosure is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0062] Unless otherwise defined, the technical and scientific terms used in the present disclosure have the same meanings as those generally understood by those skilled in the art to which the present disclosure belongs; the terms used in the present disclosure are only for the purpose of describing specific embodiments and are not intended to limit the present disclosure; the term "including" and any variations thereof in the present disclosure are intended to cover non-exclusive inclusions. Obviously, the embodiments described in the present disclosure are only part of the embodiments of the present disclosure, not all of the embodiments. Based on the embodiments in the present disclosure, all other embodiments obtained by ordinary technicians in the field without making creative efforts are within the scope of protection of the present disclosure.

[0063] Reference to "embodiments" in this disclosure means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the disclosure. The appearance of the phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0064] Figure 1 is a flowchart of a social recommendation method based on heterogeneous hypergraph and meta-path shown in an embodiment of the present disclosure. The method of the present disclosure can be executed by a terminal device, wherein the terminal device can be a server, a desktop computer, a laptop computer, etc. The embodiment of the present disclosure does not limit the type of the terminal device. Figure 1 As shown, the method includes:

[0065] Step 1: Mining implicit relationships F between users U Implicit relationship F between items I , enrich the original data and improve the recommendation accuracy;

[0066] Step 2, determining the user trust matrix B according to the 0-1 trust relationship between users and the topological structure matrix T;

[0067] Step 3: Divide user communities. Different user communities have different interest distributions. A hypergraph is constructed based on the divided user communities. A user-item interaction graph G is constructed based on the user historical interaction data. R =<U∪V,A> , where U and V represent the user set and item set respectively, A is transformed from the user-item interaction matrix R, where M represents the number of users and N represents the number of items:

[0068]

[0069] Step 4: Construct the user social network graph G based on the user trust level B S =<U,B> , S represents the user's social connection

[0070] Step 5: Transform the user-item interaction graph G R With the user social network graph G S The fusion is a fusion graph F;

[0071] Step 6: Randomly select the most influential K users on the trust matrix B;

[0072] Step 7: Randomly walk on the fusion graph F to obtain the correlation coefficient matrix S between the most influential K users and their neighbor nodes = [s i,k ] (M+N)×K ;

[0073] Step 8, divide the user community according to the correlation coefficient matrix S and construct the hypergraph matrix H;

[0074] Step 9: Pre-train the hypergraph matrix H to learn the high-order information of the nodes. After pre-training, the initial embedding representation P of the user and the item is obtained. M×d and Q N×d , d represents the embedding dimension;

[0075] Step 10: F U and F I Integrate into the fusion graph F to generate a heterogeneous graph G. Meta-paths based on user nodes and item nodes are defined in G. For user nodes, the user in-degree features are defined based on the in-degree features and out-degree features. The user in-degree features are divided into 5 types of meta-paths, and the user out-degree features are also divided into 5 types of meta-paths. For item nodes, their meta-paths are divided into 3 types. Since some nodes have many meta-paths, the dropnode mechanism is used to filter them.

[0076] Step 11: Send the five types of meta-paths of the user's in-degree features into the meta-path-based graph attention network to obtain five embeddings with different meanings ( and );

[0077] Step 12: Use a multi-layer perceptron to fuse these five embeddings with different meanings to obtain the final in-degree feature representation of user i.

[0078] Step 13: Get the out-degree feature representation of user i

[0079] Step 14: Use a multi-layer perceptron to fuse the in-degree feature representation of user i And out-degree feature representation Get the final feature representation p of user i i ;

[0080] Step 15: Send the three types of meta-paths of item j’s features into the meta-path-based graph attention network to obtain three embeddings (q j1 ,q j2 and q j3 );

[0081] Step 16: Use a multi-layer perceptron to fuse these three embeddings with different meanings to obtain the final feature representation q of item j. j ;

[0082] In a possible implementation, since the same community is more likely to express the same interest preference, the Centerloss loss term is used for constraint;

[0083] Step 17: Predict user i’s rating of item j through a multi-layer perceptron

[0084]

[0085] Step 18: Since dropnode loses the stability of prediction, a linear model is used to predict a stable score.

[0086] The final rating of user i for item j is expressed as Generate a list of recommendations.

[0087] The present invention utilizes hypergraph convolution to learn complex high-order information in the pre-training stage, and injects this high-order information into the meta-path-based graph attention network to further model users and items to obtain the final embedding representation. In the meta-path-based graph attention network, user features are divided into in-degree features and out-degree features, which will further improve the recommendation effect and use social networks to provide users with more accurate recommendation results.

[0088] According to one aspect of an embodiment of the present disclosure, a social recommendation device based on heterogeneous hypergraphs and meta-paths is provided, the device comprising:

[0089] Mining module, used to mine implicit relationships between users F U Implicit relationship F between items I ;

[0090] A first determination module is used to determine a user trust degree matrix B according to a 0-1 trust relationship between users and a topological structure matrix T;

[0091] The partitioning module is used to divide user communities. Different user communities have different interest distributions. A hypergraph is constructed based on the divided user communities. A user-item interaction graph G is constructed based on the user's historical interaction data. R =<U∪V,A> , where U and V represent the user set and item set respectively, A is transformed from the user-item interaction matrix R, where M represents the number of users and N represents the number of items:

[0092]

[0093] The first building module is used to build a user social network graph G based on the user trust level B. S =<U,B> ;

[0094] The first fusion module is used to transform the user-item interaction graph G R With the user social network graph G S The fusion is a fusion graph F;

[0095] The first selection module is used to randomly select the most influential K users on the trust matrix B;

[0096] The second selection module is used to randomly walk on the fusion graph F to obtain the correlation coefficient matrix S = [s i,k ] (M+N)×K ;

[0097] The second construction module is used to divide the user community according to the correlation coefficient matrix S and construct the hypergraph matrix H;

[0098] The training module is used to learn the high-order information of nodes through pre-training according to the hypergraph matrix H. After pre-training, the initial embedding representation P of users and items is obtained. M×d and Q N×d , d represents the embedding dimension;

[0099] Generate a module for F U and F I Integrate into the fusion graph F to generate a graph attention network G, in which meta-paths based on user nodes and item nodes are defined. For user nodes, the user in-degree features are defined based on the in-degree features and out-degree features. The user in-degree features are divided into 5 types of meta-paths, the user out-degree features are also divided into 5 types of meta-paths, and the meta-paths of item nodes are divided into 3 types.

[0100] The second determination module is used to send the five types of meta-paths of the user's in-degree features into the meta-path-based graph attention network to obtain five embeddings with different meanings ( and );

[0101] The user in-degree feature module is used to fuse these five embeddings with different meanings through a multi-layer perceptron to obtain the final in-degree feature representation of user i.

[0102] User out-degree feature module, used to obtain the out-degree feature representation of user i

[0103] User feature representation module, used to fuse the in-degree feature representation of user i through a multi-layer perceptron And out-degree feature representation Get the final feature representation p of user i i ;

[0104] The item feature determination module is used to send the three types of meta-paths of item j features into the meta-path-based graph attention network to obtain three embeddings (q j1 ,q j2 and q j3 );

[0105] The item feature representation module is used to fuse these three embeddings with different meanings through a multi-layer perceptron to obtain the final feature representation q of item j. j ;

[0106] The rating module is used to predict the rating of user i for item j through a multi-layer perceptron.

[0107]

[0108] The recommendation list determination module is used to determine the recommended list based on multiple Create a recommendation list.

[0109] In one possible implementation, a linear model is used to predict stable ratings The final score of user i on item j is expressed as And generate a recommendation list.

[0110] In a possible implementation, a Centerloss loss term is used to constrain the same user community.

[0111] In one possible implementation, in the graph attention network, a dropnode mechanism is used to filter node meta-paths.

[0112] The description of the above-mentioned device has been explained in detail in the description of the above-mentioned method, and will not be repeated here.

[0113] Figure 2 1900 is a flow chart of a social recommendation device based on heterogeneous hypergraph and meta-path according to an embodiment of the present disclosure. For example, the device 1900 may be provided as a server. Figure 2 , the apparatus 1900 includes a processing component 1922, which further includes one or more processors, and a memory resource represented by a memory 1932 for storing instructions, such as an application, that can be executed by the processing component 1922. The application stored in the memory 1932 may include one or more modules, each corresponding to a set of instructions. In addition, the processing component 1922 is configured to execute instructions to perform the above method.

[0114] The device 1900 may also include a power supply component 1926 configured to perform power management of the device 1900, a wired or wireless network interface 1950 configured to connect the device 1900 to a network, and an input / output (I / O) interface 1958. The device 1900 may operate based on an operating system stored in the memory 1932, such as Windows Server™, MacOS X™, Unix™, Linux™, FreeBSD™, or the like.

[0115] In an exemplary embodiment, a non-volatile computer-readable storage medium is also provided, such as a memory 1932 including computer program instructions, which can be executed by the processing component 1922 of the device 1900 to perform the above method.

[0116] The present disclosure may be a system, a method and / or a computer program product. The computer program product may include a computer-readable storage medium carrying computer-readable program instructions for causing a processor to implement various aspects of the present disclosure.

[0117] A computer-readable storage medium may be a tangible device that can hold and store instructions used by an instruction execution device. A computer-readable storage medium may be, for example, but not limited to, an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the foregoing. More specific examples of computer-readable storage media (a non-exhaustive list) include: 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), a static random access memory (SRAM), a portable compact disk read-only memory (CD-ROM), a digital versatile disk (DVD), a memory stick, a floppy disk, a mechanical encoding device, such as a punch card or a raised structure in a groove on which instructions are stored, and any suitable combination of the foregoing. As used herein, a computer-readable storage medium is not to be interpreted as a transient signal per se, such as a radio wave or other freely propagating electromagnetic wave, an electromagnetic wave propagating through a waveguide or other transmission medium (e.g., a light pulse through a fiber optic cable), or an electrical signal transmitted through a wire.

[0118] The computer-readable program instructions described herein can be downloaded from a computer-readable storage medium to each computing / processing device, or downloaded to an external computer or external storage device via a network, such as the Internet, a local area network, a wide area network, and / or a wireless network. The network can include copper transmission cables, optical fiber transmissions, wireless transmissions, routers, firewalls, switches, gateway computers, and / or edge servers. The network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards the computer-readable program instructions for storage in the computer-readable storage medium in each computing / processing device.

[0119] The computer program instructions for performing the operation of the present disclosure may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-related instructions, microcode, firmware instructions, state setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages, such as Smalltalk, C++, etc., and conventional procedural programming languages, such as "C" language or similar programming languages. Computer-readable program instructions may be executed completely on a user's computer, partially on a user's computer, as an independent software package, partially on a user's computer, partially on a remote computer, or completely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., using an Internet service provider to connect via the Internet). In some embodiments, an electronic circuit, such as a programmable logic circuit, a field programmable gate array (FPGA), or a programmable logic array (PLA), may be customized by utilizing the state information of the computer-readable program instructions, and the electronic circuit may execute the computer-readable program instructions, thereby realizing various aspects of the present disclosure.

[0120] Various aspects of the present disclosure are described herein with reference to the flowcharts and / or block diagrams of the methods, devices (systems) and computer program products according to the embodiments of the present disclosure. It should be understood that each box in the flowchart and / or block diagram and the combination of each box in the flowchart and / or block diagram can be implemented by computer-readable program instructions.

[0121] These computer-readable program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, thereby producing a machine, so that when these instructions are executed by the processor of the computer or other programmable data processing device, a device that implements the functions / actions specified in one or more boxes in the flowchart and / or block diagram is generated. These computer-readable program instructions can also be stored in a computer-readable storage medium, and these instructions cause the computer, programmable data processing device, and / or other equipment to work in a specific manner, so that the computer-readable medium storing the instructions includes a manufactured product, which includes instructions for implementing various aspects of the functions / actions specified in one or more boxes in the flowchart and / or block diagram.

[0122] Computer-readable program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other device so that a series of operating steps are performed on the computer, other programmable data processing apparatus, or other device to produce a computer-implemented process, thereby causing the instructions executed on the computer, other programmable data processing apparatus, or other device to implement the functions / actions specified in one or more boxes in the flowchart and / or block diagram.

[0123] The flow chart and block diagram in the accompanying drawings show the possible architecture, function and operation of the system, method and computer program product according to multiple embodiments of the present disclosure. In this regard, each square box in the flow chart or block diagram can represent a part of a module, program segment or instruction, and a part of the module, program segment or instruction includes one or more executable instructions for realizing the specified logical function. In some alternative implementations, the function marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two continuous square boxes can actually be executed substantially in parallel, and they can sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs the specified function or action, or can be implemented with a combination of special hardware and computer instructions.

[0124] The embodiments of the present disclosure have been described above, and the above description is exemplary, not exhaustive, and is not limited to the disclosed embodiments. Many modifications and changes will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The selection of terms used herein is intended to best explain the principles of the embodiments, practical applications, or improvements to the technology in the market, or to enable other persons of ordinary skill in the art to understand the embodiments disclosed herein.

Claims

1. A social recommendation method based on heterogeneous hypergraph and meta-path, characterized in that: The method comprises: Step 1: Mining implicit relationships F between users U Implicit relationship F between items I ; Step 2, determining the user trust matrix B according to the 0-1 trust relationship between users and the topological structure matrix T; Step 3: Divide user communities. Different user communities have different interest distributions. A hypergraph is constructed based on the divided user communities. A user-item interaction graph G is constructed based on the user historical interaction data. R =<U∪V,A> , where U and V represent the user set and item set respectively, A is transformed from the user-item interaction matrix R, where M represents the number of users and N represents the number of items: Step 4: Construct the user social network graph G based on the user trust level B S =<U,B> ; Step 5: Transform the user-item interaction graph G R With the user social network graph G S The fusion is a fusion graph F; Step 6: Randomly select the most influential K users on the trust matrix B; Step 7: Randomly walk on the fusion graph F to obtain the correlation coefficient matrix S = [s i,k ] (M+N)×K ; Step 8, divide the user community according to the correlation coefficient matrix S and construct the hypergraph matrix H; Step 9: Pre-train the hypergraph matrix H to learn the high-order information of the nodes. After pre-training, the initial embedding representation P of the user and the item is obtained. M×d and Q N×d , d represents the embedding dimension; Step 10: F U and F I Integrate into the fusion graph F to generate a graph attention network G, in which meta-paths based on user nodes and item nodes are defined. For user nodes, the user in-degree features are defined based on the in-degree features and out-degree features. The user in-degree features are divided into 5 types of meta-paths, the user out-degree features are also divided into 5 types of meta-paths, and the meta-paths of item nodes are divided into 3 types. Step 11: Send the five types of meta-paths of the user's in-degree features into the meta-path-based graph attention network to obtain five embeddings with different meanings ( and ); Step 12: Use a multi-layer perceptron to fuse these five embeddings with different meanings to obtain the final in-degree feature representation of user i. Step 13: Get the out-degree feature representation of user i Step 14: Use a multi-layer perceptron to fuse the in-degree feature representation of user i And out-degree feature representation Get the final feature representation p of user i i ; Step 15: Send the three types of meta-paths of item j’s features into the meta-path-based graph attention network to obtain three embeddings (q j1 ,q j2 and q j3 ); Step 16: Use a multi-layer perceptron to fuse these three embeddings with different meanings to obtain the final feature representation q of item j. j ; Step 17: Predict user i’s rating of item j through a multi-layer perceptron Step 18: Based on multiple Create a recommendation list.

2. The method according to claim 1, characterized in that Predicting stable ratings using linear models The final score of user i on item j is expressed as And generate a recommendation list.

3. The method according to claim 1, characterized in that Centerloss is used to constrain the same user community.

4. The method according to claim 1, characterized in that: In the graph attention network, the dropnode mechanism is used to filter the node meta-path.

5. A social recommendation device based on heterogeneous hypergraph and meta-path, characterized in that: The device comprises: Mining module, used to mine implicit relationships between users F U Implicit relationship F between items I ; A first determination module is used to determine a user trust degree matrix B according to a 0-1 trust relationship between users and a topological structure matrix T; The partitioning module is used to partition user communities. Different user communities have different interest distributions. A hypergraph is constructed based on the partitioned user communities, and a user-item interaction graph is constructed based on the user's historical interaction data. R =<U∪V,A> , where U and V represent the user set and item set respectively, A is transformed from the user-item interaction matrix R, where M represents the number of users and N represents the number of items: The first building module is used to build a user social network graph G based on the user trust level B. s =<U,B> ; The first fusion module is used to transform the user-item interaction graph G R With the user social network graph G S The fusion is a fusion graph F; The first selection module is used to randomly select the most influential K users on the trust matrix B; The second selection module is used to randomly walk on the fusion graph F to obtain the correlation coefficient matrix S = [s i,k ] (M+N)×K ; The second construction module is used to divide the user community according to the correlation coefficient matrix S and construct the hypergraph matrix H; The training module is used to learn the high-order information of nodes through pre-training according to the hypergraph matrix H. After pre-training, the initial embedding representation P of users and items is obtained. M×d and Q N×d , d represents the embedding dimension; Generate a module for F U and F I Integrate into the fusion graph F to generate a graph attention network G, in which meta-paths based on user nodes and item nodes are defined. For user nodes, the user in-degree features are defined based on the in-degree features and out-degree features. The user in-degree features are divided into 5 types of meta-paths, the user out-degree features are also divided into 5 types of meta-paths, and the meta-paths of item nodes are divided into 3 types. The second determination module is used to send the five types of meta-paths of the user's in-degree features into the meta-path-based graph attention network to obtain five embeddings with different meanings ( and ); The user in-degree feature module is used to fuse these five embeddings with different meanings through a multi-layer perceptron to obtain the final in-degree feature representation of user i. User out-degree feature module, used to obtain the out-degree feature representation of user i User feature representation module, used to fuse the in-degree feature representation of user i through a multi-layer perceptron And out-degree feature representation Get the final feature representation p of user i i ; The item feature determination module is used to send the three types of meta-paths of item j features into the meta-path-based graph attention network to obtain three embeddings (q j1 ,q j2 and q j3 ); The item feature representation module is used to fuse these three embeddings with different meanings through a multi-layer perceptron to obtain the final feature representation q of item j. j ; The rating module is used to predict the rating of user i for item j through a multi-layer perceptron. The recommendation list determination module is used to determine the recommendation list based on multiple Create a recommendation list.

6. The device according to claim 1, characterized in that Predicting stable ratings using linear models The final score of user i on item j is expressed as And generate a recommendation list.

7. The device according to claim 1, characterized in that Centerloss is used to constrain the same user community.

8. The device according to claim 1, characterized in that In the graph attention network, the dropnode mechanism is used to filter the node meta-path.

9. A social recommendation device based on heterogeneous hypergraph and meta-path, characterized in that: The device comprises: processor; a memory for storing processor-executable instructions; The processor is configured to execute the method according to any one of claims 1 to 4.

10. A non-volatile computer-readable storage medium having computer program instructions stored thereon, characterized in that: When the computer program instructions are executed by a processor, the method according to any one of claims 1 to 4 is implemented.

Images