Community detection method, device and equipment in network

Abstract

The invention provides a community detection method, device and equipment in a network. The method comprises the following steps: shrinking different connecting edges in a target network to generate different new networks; determining the connection edge centrality of the target network based on the difference between the target network and the new network; determining weak connection edges and a plurality of communities in the target network based on the connection edge centrality identification, wherein points in the weak connection edges are located at the edges of the communities; clustering the nodes in the adjacent communities by taking the node in the weak connection edge as a starting point, updating based on a clustering result, and determining the edge of the community; and reallocating and adjusting the nodes in the community based on the edge connection condition of each node so as to obtain a detection result corresponding to the community in the target network. According to the community detection method in the network, the weak connection and the connection edge centrality in the network can be identified and detected more quickly and accurately, and the community in the network can be identified quickly based on the weak connection and the connection edge centrality.

Description

Technical Field

[0001] The present invention belongs to the technical field of network communities, and in particular relates to a community detection method, device and equipment within a network. Background Art

[0002] The importance of ties has been a focus of research in various network analysis tasks over the past few decades, including backbone structure identification, epidemic control, and network collapse. Recent studies have revealed the impact of different types of ties on the network and emphasized the significance of weak ties. These weak ties were originally considered to be channels for propagating useful information to groups with longer paths, considering the length of time, intimacy, and reciprocity between node pairs. Recently, the definition of weak ties has been expanded to cross-community ties, which are often considered to be bridges that facilitate the flow of information between different communities. Although many community detection algorithms have been proposed to identify weak ties from the topological structure, in many cases, cross-community ties can be transformed into intra-community ties due to the ambiguous or overlapping community structure of the network. In this case, quantifying weak ties will be more convincing. Therefore, it is crucial to accurately assess the ties centrality, especially in the context of cross-community ties.

[0003] A large number of metrics have been developed to evaluate the role of weak connections in facilitating communication between different communities, such as topological overlap, bridging, diffusion importance, betweenness, etc. These metrics aim to quantify the role of weak connections in maintaining local or global connectivity from the perspectives of neighborhood, connection group, node information diffusion, and shortest path. In addition, evaluating the importance of edges in maintaining global connectivity also involves the topological difference between the original network and the network after shrinking the edges. The shrunken network is generated by merging pairs of nodes into a single node connected to all its neighbors. Many methods have been proposed to quantify this difference, among which node distance distribution has been shown to be very effective in identifying subtle topological changes. The network node distance distribution provides a detailed and compact representation of the network topology and has inspired various research works, including diffusion capacity, link prediction, community fragility, etc. However, this difference measurement fails to describe in detail the impact of edges on maintaining the high-order neighborhood of nodes, that is, the hierarchical structure.

[0004] In order to evaluate the effectiveness of topology-based edge centrality in identifying weak ties, cross-community edges based on the Louvain algorithm are regarded as real weak ties. Edge-based network collapse is introduced to examine the impact of weak ties on global connectivity. Classification accuracy and average maximum connected subgraph size indicate the importance of edge centrality in capturing weak ties and maintaining global connectivity, respectively. Some community detection algorithms are based on the relationship between edge centrality and community structure. For example, the Girvan-Newman (GN) method detects communities by recalculating the edge betweenness between node pairs, verifying the effectiveness of betweenness in separating node pairs and quantifying weak ties. However, the GN method tends to prioritize the edge order of separated nodes, so it cannot directly capture the degree to which betweenness describes community structure. In addition, the priority of weak ties is not fully utilized when dividing communities. For example, in networks with significant community structure, the edge centrality distribution of weak ties and intra-community edges shows obvious differences. An effective threshold helps to identify most weak ties, thereby helping to determine the number of communities and minimize the bias caused by the initial division. Therefore, a community detection framework based on edge ranking is necessary to explore the community structure information contained in edge centrality. Summary of the invention

[0005] The technical problem to be solved by the present invention is to provide a community detection method, device and equipment within a network that can more quickly and accurately identify weak connections and edge centrality in the network, and quickly identify communities in the network based on the weak connections and edge centrality.

[0006] The present invention includes a community detection method in a network, comprising:

[0007] Contract different edges in the target network to generate different new networks;

[0008] Determining the edge centrality of the target network based on the difference between the target network and the new network;

[0009] Based on the edge centrality identification, weak edges and multiple communities in the target network are determined, and points in the weak edges are located at the edge of the community;

[0010] Aggregating nodes in adjacent communities using the nodes in the weak links as starting points, and updating and determining the edges of the communities based on the aggregation results;

[0011] The nodes in the community are reallocated and adjusted based on the edge connection status of each of the nodes, thereby obtaining a detection result corresponding to the community in the target network.

[0012] In some embodiments, after generating different new networks, the method further includes:

[0013] Determining a node hierarchy of the target network, wherein the node hierarchy is the number of node neighbors at different topological distances obtained by a breadth-first search method;

[0014] A network hierarchy of the target network is determined based on the node hierarchy, where the network hierarchy includes network hierarchies corresponding to different edges.

[0015] In some embodiments, the node hierarchy is represented by a node spindle feature, wherein the node spindle feature It is expressed as the number of nodes away from node i, and the formula is as follows:

[0016]

[0017] in Representation level L i The node ratio in, the root node is regarded as the 0th layer;

[0018] The network hierarchy is formed by aggregating the hierarchical structures of all nodes.

[0019] In some embodiments, the method further comprises:

[0020] The difference between the target network before shrinking the edges and the new network formed after shrinking the edges is evaluated by the JS divergence of the network hierarchy of the shrunken edges.

[0021] In some embodiments, determining the edge centrality of the target network based on the difference between the target network and the new network includes:

[0022] After determining the target edge to be shrunk, in the network hierarchy of the target edge, except for the nodes and nodes constituting the target edge, the node information of the remaining nodes in the original network and the new network;

[0023] Calculate the sum of JS divergences of corresponding nodes based on the node information;

[0024] Determining the difference between the original network and the new network caused by shrinking the target edge based on the sum of the JS divergences;

[0025] An edge centrality corresponding to the target edge is determined based on the difference.

[0026] In some embodiments, the aggregating nodes in adjacent communities using the nodes in the weak links as starting points, and updating and determining the edges of the communities based on the aggregation results, includes:

[0027] Aggregating nodes in adjacent communities using the nodes in the weak links as starting points to obtain aggregation results;

[0028] If the aggregation result indicates that an intra-community node connection is established between the node in the weak connection and a node in the community, the node in the weak connection is divided into the community.

[0029] In some embodiments, the reallocation and adjustment of nodes in the community based on the edge connection status of each node includes:

[0030] If the number of connections between the node in the weak connection and the nodes in the first community is greater than the number of connections between the node in the weak connection and the nodes in the second community, the node in the weak connection is divided into the first community;

[0031] If the number of links between the nodes in the weak links and the nodes in the first community and the second community is equal, the average link centrality of the first community and the second community is calculated respectively, and the square with the smaller average link centrality is determined;

[0032] The nodes in the weak links are divided into communities corresponding to the smaller average link centrality.

[0033] In some embodiments, the reallocation and adjustment of nodes in the community based on the edge connection status of each node includes:

[0034] Calculate and determine the redundancy of the nodes in the weak links in different communities, where the redundancy of the nodes in the community represents the number of links that are redundant compared to maintaining the hierarchical relationship of the first-order and second-order neighbors in the community;

[0035] The nodes in the weak connection are divided into a target community with maximum redundancy.

[0036] Another embodiment of the present invention also provides a community detection device in a network, including:

[0037] The contraction module is used to contract different edges in the target network to generate different new networks;

[0038] A first determination module, configured to determine the edge centrality of the target network according to the difference between the target network and the new network;

[0039] A second determination module is used to identify and determine weak edges and multiple communities in the target network according to the edge centrality, wherein the points in the weak edges are located at the edge of the community;

[0040] An aggregation module, used to aggregate nodes in adjacent communities with the nodes in the weak links as starting points, and update and determine the edges of the communities based on the aggregation results;

[0041] The adjustment module is used to reallocate and adjust the nodes in the community according to the edge connection status of each node, so as to obtain the detection result of the community in the target network.

[0042] Another embodiment of the present invention further provides an electronic device, comprising:

[0043] one or more processors;

[0044] a memory configured to store one or more programs;

[0045] When the one or more programs are executed by the one or more processors, the one or more processors implement the community detection method within a network as described in any one of the embodiments above.

[0046] The beneficial effects of the present invention include generating different new networks by shrinking different edges in the target network; determining the edge centrality of the target network based on the difference between the target network and the new network; identifying and determining weak edges and multiple communities in the target network based on the edge centrality, wherein the points in the weak edges are located at the edge of the community; aggregating the nodes in the adjacent communities with the nodes in the weak edges as the starting point, and updating and determining the edge of the community based on the aggregation results; redistributing and adjusting the nodes in the community based on the edge conditions of each of the nodes, thereby obtaining the detection results of the communities in the target network. The above method of this embodiment can accurately identify the weak connections in the target network, and determine the edge centrality of the edges between different nodes based on the weak connection evaluation, and then obtain each community in the target network based on the rapid detection of the weak connection and edge centrality, so that not only the detection and identification of the community can be realized, but also the network connectivity can be accurately characterized and the relationship between the weak connection and the community can be formalized.

[0047] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description, claims, and drawings.

[0048] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0050] Figure 1Schematic diagram of the process of a community detection method in a network according to an embodiment of the present invention.

[0051] Figure 2 FIG. 4 is a flow chart of a community detection method in a network according to another embodiment of the present invention.

[0052] Figure 3 FIG. 4 is a flow chart of a community detection method in a network according to another embodiment of the present invention.

[0053] Figure 4 4 is a structural block diagram of a community detection device in a network according to an embodiment of the present invention. DETAILED DESCRIPTION

[0054] The specific embodiments of the present invention are described in detail below with reference to the accompanying drawings, but are not intended to limit the present invention.

[0055] It should be understood that various modifications may be made to the embodiments disclosed herein. Therefore, the following description should not be considered as limiting, but merely as an example of an embodiment. Other modifications within the scope of the invention will occur to those skilled in the art.

[0056] The accompanying drawings, which are incorporated in and constitute a part of the specification, illustrate embodiments of the invention and, together with the general description of the invention given above and the detailed description of the embodiments given below, serve to explain the principles of the invention.

[0057] These and other characteristics of the invention will become apparent from the following description of a preferred form of embodiment given as a non-limiting example, with reference to the accompanying drawings.

[0058] It should also be understood that, although the invention has been described with reference to some specific examples, those skilled in the art will be able to realize many other equivalent forms of the invention that have the characteristics recited in the claims and are therefore within the scope of protection defined thereby.

[0059] The above and other aspects, features and advantages of the present invention will become more apparent from the following detailed description when taken in conjunction with the accompanying drawings.

[0060] Specific embodiments of the present invention are described hereinafter with reference to the accompanying drawings; however, it should be understood that the disclosed embodiments are merely examples of the present invention, which may be implemented in a variety of ways. Well-known and / or repeated functions and structures are not described in detail to avoid unnecessary or redundant details that obscure the present invention. Therefore, the specific structural and functional details disclosed herein are not intended to be limiting, but merely serve as a basis and representative basis for the claims to teach those skilled in the art to use the present invention in a variety of ways with substantially any suitable detailed structure.

[0061] This specification may use the phrases "in one embodiment," "in another embodiment," "in a further embodiment," or "in other embodiments," which may all refer to one or more of the same or different embodiments according to the present invention.

[0062] The embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0063] like Figure 1 As shown, an embodiment of the present invention provides a community detection method in a network, including:

[0064] S1: Contract different edges in the target network to generate different new networks;

[0065] S2: determining the edge centrality of the target network based on the difference between the target network and the new network;

[0066] S3: Identify weak edges and multiple communities in the target network based on the edge centrality, where points in the weak edges are located at the edge of the community;

[0067] S4: Aggregate the nodes in the adjacent communities with the nodes in the weak link as the starting point, and update and determine the edge of the community based on the aggregation result;

[0068] S5: reallocate and adjust the nodes in the community based on the edge connection status of each of the nodes, and then obtain a detection result corresponding to the community in the target network.

[0069] Based on the above content, it can be seen that this embodiment proposes a new community detection framework, which generates different new networks by shrinking different edges in the target network; determines the edge centrality of the target network based on the difference between the target network and the new network; identifies weak edges and multiple communities in the target network based on the edge centrality, and the points in the weak edges are located at the edge of the community; aggregates the nodes in the adjacent communities with the nodes in the weak edges as the starting point, and updates and determines the edge of the community based on the aggregation results; redistributes and adjusts the nodes in the community based on the edge situation of each node, and then obtains the detection result of the community in the target network. Based on the above method of this embodiment, the weak connection in the target network can be accurately identified, and the edge centrality of the edges between different nodes can be determined based on the weak connection evaluation, and then the various communities in the target network can be obtained based on the rapid detection of the weak connection and edge centrality, so that not only the detection and identification of the community can be realized, but also the network connectivity can be accurately characterized, the relationship between weak connections and communities can be formalized, and the importance of edge centrality in capturing community structure can be evaluated.

[0070] In one embodiment, if Figure 2As shown, after generating different new networks, the method further includes:

[0071] S6: Determine a node hierarchy of the target network, wherein the node hierarchy is the number of node neighbors at different topological distances obtained by a breadth-first search method;

[0072] S7: Determine a network hierarchy of the target network based on the node hierarchy, where the network hierarchy includes network hierarchies corresponding to different edges.

[0073] Specifically, unlike most traditional definitions, in this embodiment, the node hierarchy refers to the number of node neighbors at different topological distances obtained by the breadth-first search (BFS) method. This framework does not need to consider the specific connections between nodes when describing the network topology from a meso-level perspective, while retaining a larger hierarchical relationship. In the target network, the hierarchy shows a trend of gradually shrinking after increasing to a peak, similar to a "spindle" shape, where the node at level l represents a neighbor node that is l away from the root node i. Therefore, in this embodiment, the node hierarchy is represented by a node spindle feature, and the node spindle feature It is expressed as the number of nodes away from node i, and the formula is as follows:

[0074]

[0075] in Representation level L i The node ratio in, the root node is regarded as the 0th layer;

[0076] The network hierarchy is formed by aggregating the hierarchical structures of all nodes, which is expressed as follows:

[0077]

[0078] The network hierarchy can clearly reveal characteristics such as degree distribution, diameter, and average shortest distance length of the network. In addition, synthetic networks with different generation mechanisms show different hierarchical patterns, while different configurations of the same network model show obvious similarities.

[0079] In another embodiment, the method further comprises:

[0080] S8: The difference between the target network before shrinking the edges and the new network formed after shrinking the edges is evaluated by the JS divergence of the network hierarchy of the shrunken edges.

[0081] Among them, Figure 3 As shown, determining the edge centrality of the target network based on the difference between the target network and the new network includes:

[0082] S9: after determining the target edge to be shrunk, in the network hierarchy of the target edge, except for the nodes and nodes constituting the target edge, the node information of the remaining nodes in the original network and the new network;

[0083] S10: Calculate the sum of JS divergences of corresponding nodes based on the node information;

[0084] S11: determining the difference between the original network and the new network caused by shrinking the target edge based on the sum of the JS divergences;

[0085] S12: Determine the edge centrality corresponding to the target edge based on the difference.

[0086] Specifically, in this embodiment, the dissimilarity between the original network and the new network is evaluated by the Jensen-Shannon (JS) divergence of the network hierarchy based on the edge centrality of the hierarchical structure. The target network is an undirected, unweighted network G with N nodes and E edges, and the edge to be shrunk is the target edge (specific edge) e ij For example, in the network hierarchy of the target edge, the N-nodes other than nodes i and j are represented as follows in the initial network and the new network:

[0087]

[0088] Target edge ij The edge centrality is expressed as the sum of the JS divergence values ​​of N-2 nodes:

[0089]

[0090] Where CN(i,j) represents the number of common neighbors between nodes i and j. is the JS divergence of the spindle vector of node k:

[0091]

[0092] where KL(P,Q) is the KL divergence of distributions P and Q:

[0093] KL(P,Q)=∑P(x)log(P(x) / Q(x))

[0094] Compared with nodes within a community, merging nodes connected by weak edges will significantly change the topology of the network, and thus change the network hierarchy to a greater extent.

[0095] Furthermore, in this embodiment, the nodes connected by weak links are located at the edge of the community, and the nodes indirectly connected to these edges through multiple links within the community are regarded as internal nodes of the corresponding community. In this embodiment, the nodes in the adjacent communities are aggregated with the nodes in the weak links as the starting points, and the edge of the community is updated and determined based on the aggregation results, including:

[0096] S13: Aggregating nodes in adjacent communities using the nodes in the weak link as starting points to obtain aggregation results;

[0097] S14: If the aggregation result indicates that an intra-community node connection is established between the node in the weak connection and a node in the community, the node in the weak connection is divided into the community.

[0098] S15: Nodes connected by weak edges are usually located at the edges of communities, while nodes indirectly connected to these edges through multiple edges within the community are usually regarded as internal nodes of the corresponding community.

[0099] For example, based on the threshold k (community division number threshold) of the edge centrality metric M, the edges are classified into internal edges and weak edges within the community. This classification helps to identify the edge of the community through weakly connected nodes. The model in this embodiment performs node aggregation on the adjacent communities of the node starting from the edge node (i.e., the node in the weak connection), which is equivalent to starting with a preliminary detection of several candidate communities defined by the nodes in the weak connection, and performing node aggregation. If the aggregation result characterizes that the edge node is connected through an internal edge in the community, that is, when an internal edge is established between the edge node and a node in a community, the edge node is merged into the community, so that the community edge can be determined more accurately and the number of communities can be reduced. For example, in a simplified network, three weak edges involve four edge nodes. If an edge node and a node in a community are connected to each other through an edge in a community, it indicates that the edge node belongs to the community.

[0100] In another embodiment, the reallocation and adjustment of nodes in the community based on the edge connection status of each node includes:

[0101] S16: If the number of connections between the nodes in the weak connection and the nodes in the first community is greater than the number of connections between the nodes in the second community, the nodes in the weak connection are divided into the first community;

[0102] S17: If the number of links between the nodes in the weak links and the nodes in the first community and the second community is equal, respectively calculate the average link centrality of the first community and the second community, and determine the square with the smaller average link centrality;

[0103] S18: Divide the nodes in the weak links into communities corresponding to the smaller average link centrality.

[0104] This embodiment is used to implement the specific allocation of internal nodes. Specifically, although many nodes are often connected to multiple communities through intra-community edges, it is difficult to achieve accurate node allocation assuming that the community is composed of internal nodes with the most intra-community edges or the lowest average edge centrality. In order to ensure that the node allocation is more accurate and has a wider scope of application, the solution adopted in this embodiment is that if the edge node i has more edges in the first community than in the second community, it is allocated to the first community. In the case where the intra-community edges between the edge node i and the nodes in the first community and the second community are equal, the node i is allocated to the community with a lower average edge centrality. The method in this embodiment can avoid the problem of recalculation of edge centrality that is common in split community detection methods, and maximize the correlation between weak connections and community boundaries to model the community detection problem as a problem of edge centrality ranking and partition thresholds.

[0105] Furthermore, the redistribution and adjustment of nodes in the community based on the edge connection of each node includes:

[0106] S19: Calculate and determine the redundancy of the nodes in the weak links in different communities, where the redundancy of the nodes in the community represents the number of links that are redundant compared to maintaining the hierarchical relationship of the first-order and second-order neighbors in the community;

[0107] S20: Divide the nodes in the weak connection into a target community with maximum redundancy.

[0108] The method in this embodiment is used to implement the fine-tuning of the partition of edge nodes. Specifically, considering that the association between nodes and communities changes continuously during the node allocation process, this embodiment chooses to evaluate the partition of each node according to its redundancy after the aggregation process. It is assumed that the best partition is the community consisting of nodes with the largest redundancy. The redundancy of edge node i in community C is It represents the number of redundant edges compared to maintaining the hierarchical relationship of first-order and second-order neighbors in community C, expressed as:

[0109] in and denote the number of first-order and second-order neighbors of node i in community C, respectively. represents the number of edges connecting these neighbors. It means that the redundancy of node i in community C is relatively high. According to the above adjustment principle, node i is preferentially assigned to the community with the largest redundancy.

[0110] like Figure 4As shown, another embodiment of the present invention also provides a community detection device 100 in a network, including:

[0111] The contraction module is used to contract different edges in the target network to generate different new networks;

[0112] A first determination module, configured to determine the edge centrality of the target network according to the difference between the target network and the new network;

[0113] A second determination module is used to identify and determine weak edges and multiple communities in the target network according to the edge centrality, wherein the points in the weak edges are located at the edge of the community;

[0114] An aggregation module, used to aggregate nodes in adjacent communities with the nodes in the weak links as starting points, and update and determine the edges of the communities based on the aggregation results;

[0115] The adjustment module is used to reallocate and adjust the nodes in the community according to the edge connection status of each node, so as to obtain the detection result of the community in the target network.

[0116] In one embodiment, the device further comprises:

[0117] A third determination module is used to determine a node hierarchy of the target network, wherein the node hierarchy is the number of node neighbors at different topological distances obtained by a breadth-first search method;

[0118] A fourth determination module is used to determine a network hierarchy of the target network according to the node hierarchy, wherein the network hierarchy includes network hierarchies corresponding to different edges.

[0119] In one embodiment, the node hierarchy is represented by a node spindle feature. It is expressed as the number of nodes away from node i, and the formula is as follows:

[0120]

[0121] in Representation level L i The node ratio in, the root node is regarded as the 0th layer;

[0122] The network hierarchy is formed by aggregating the hierarchical structures of all nodes.

[0123] In one embodiment, the device further comprises:

[0124] An evaluation module is used to evaluate the difference between the target network before shrinking the edges and the new network formed after shrinking the edges through the JS divergence of the network hierarchy of the shrunken edges.

[0125] In one embodiment, determining the edge centrality of the target network based on the difference between the target network and the new network includes:

[0126] After determining the target edge to be shrunk, in the network hierarchy of the target edge, except for the nodes and nodes constituting the target edge, the node information of the remaining nodes in the original network and the new network;

[0127] Calculate the sum of JS divergences of corresponding nodes based on the node information;

[0128] Determining the difference between the original network and the new network caused by shrinking the target edge based on the sum of the JS divergences;

[0129] An edge centrality corresponding to the target edge is determined based on the difference.

[0130] In one embodiment, aggregating nodes in adjacent communities using the nodes in the weak links as starting points, and updating and determining the edges of the communities based on the aggregation results, includes:

[0131] Aggregating nodes in adjacent communities using the nodes in the weak links as starting points to obtain aggregation results;

[0132] If the aggregation result indicates that an intra-community node connection is established between the node in the weak connection and a node in the community, the node in the weak connection is divided into the community.

[0133] In one embodiment, the reallocation and adjustment of nodes in the community based on the edge connection status of each node includes:

[0134] If the number of connections between the node in the weak connection and the nodes in the first community is greater than the number of connections between the node in the weak connection and the nodes in the second community, the node in the weak connection is divided into the first community;

[0135] If the number of links between the nodes in the weak links and the nodes in the first community and the second community is equal, the average link centrality of the first community and the second community is calculated respectively, and the square with the smaller average link centrality is determined;

[0136] The nodes in the weak links are divided into communities corresponding to the smaller average link centrality.

[0137] In one embodiment, the reallocation and adjustment of nodes in the community based on the edge connection status of each node includes:

[0138] Calculate and determine the redundancy of the nodes in the weak links in different communities, where the redundancy of the nodes in the community represents the number of links that are redundant compared to maintaining the hierarchical relationship of the first-order and second-order neighbors in the community;

[0139] The nodes in the weak connection are divided into a target community with maximum redundancy.

[0140] Another embodiment of the present invention further provides an electronic device, comprising:

[0141] one or more processors;

[0142] a memory configured to store one or more programs;

[0143] When the one or more programs are executed by the one or more processors, the one or more processors implement the community detection method within a network as described in any one of the embodiments above.

[0144] Furthermore, an embodiment of the present invention also provides a storage medium on which a computer program is stored, and when the program is executed by a processor, the community detection method in the network described above is implemented. It should be understood that each solution in this embodiment has the corresponding technical effect in the above method embodiment, which will not be repeated here.

[0145] Furthermore, an embodiment of the present invention also provides a computer program product, which is tangibly stored on a computer-readable medium and includes computer-readable instructions, which, when executed, enable at least one processor to perform a community detection method within a network such as in the embodiments described above.

[0146] It should be noted that the computer storage medium of 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 medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, system 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 storage medium (RAM), a read-only storage medium (ROM), an erasable programmable read-only storage medium (EPROM or flash memory), an optical fiber, a portable compact disk read-only storage medium (CD-ROM), an optical storage medium, a magnetic storage medium, 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, system 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, wherein a computer-readable program code is carried. This propagated data signal may take a variety of forms, including but not limited to an electromagnetic signal, an optical signal, or any suitable combination of the above. A computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium that can send, propagate or transmit a program configured to be used by or in conjunction with an instruction execution system, system or device. The program code contained on the computer-readable medium may be transmitted using any appropriate medium, including but not limited to: wireless, antenna, optical cable, RF, etc., or any suitable combination of the above.

[0147] In addition, it will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage and optical storage, etc.) containing computer-usable program code.

[0148] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1A process or multiple processes and / or boxes Figure 1 A system that specifies the functions of a box or multiple boxes.

[0149] These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory produce a product including an instruction system, which is implemented in the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0150] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0151] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

[0152] One or more embodiments of the present application are intended to cover all such substitutions, modifications and variations that fall within the broad scope of the present application. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of one or more embodiments of the present application should be included in the protection scope of the present application.

Claims

1. A community detection method in a network, characterized in that: include: Contract different edges in the target network to generate different new networks; Determining the edge centrality of the target network based on the difference between the target network and the new network; Based on the edge centrality identification, weak edges and multiple communities in the target network are determined, and points in the weak edges are located at the edge of the community; Aggregating nodes in adjacent communities using the nodes in the weak links as starting points, and updating and determining the edges of the communities based on the aggregation results; The nodes in the community are reallocated and adjusted based on the edge connection status of each of the nodes, thereby obtaining a detection result corresponding to the community in the target network.

2. The community detection method in a network according to claim 1, characterized in that: After generating different new networks, the method further includes: Determining a node hierarchy of the target network, wherein the node hierarchy is the number of node neighbors at different topological distances obtained by a breadth-first search method; A network hierarchy of the target network is determined based on the node hierarchy, where the network hierarchy includes network hierarchies corresponding to different edges.

3. The community detection method in a network according to claim 2, characterized in that: The node hierarchy is represented by a node spindle feature. It is expressed as the number of nodes away from node i, and the formula is as follows: in Representation level L i The node ratio in, the root node is regarded as the 0th layer; The network hierarchy is formed by aggregating the hierarchical structures of all nodes.

4. The community detection method in a network according to claim 3, characterized in that: The method further comprises: The difference between the target network before shrinking the edges and the new network formed after shrinking the edges is evaluated by the JS divergence of the network hierarchy of the shrunken edges.

5. The community detection method in a network according to claim 4, characterized in that: The determining the edge centrality of the target network based on the difference between the target network and the new network includes: After determining the target edge to be shrunk, in the network hierarchy of the target edge, except for the nodes and nodes constituting the target edge, the node information of the remaining nodes in the original network and the new network; Calculate the sum of JS divergences of corresponding nodes based on the node information; Determining the difference between the original network and the new network caused by shrinking the target edge based on the sum of the JS divergences; An edge centrality corresponding to the target edge is determined based on the difference.

6. The community detection method in a network according to claim 1, characterized in that: The step of aggregating nodes in adjacent communities using the nodes in the weak links as starting points, and updating and determining the edges of the communities based on the aggregation results, includes: Aggregating nodes in adjacent communities using the nodes in the weak links as starting points to obtain aggregation results; If the aggregation result indicates that an intra-community node connection is established between the node in the weak connection and a node in the community, the node in the weak connection is divided into the community.

7. The method for detecting a community in a network according to claim 1, characterized in that: The reallocation and adjustment of nodes in the community based on the edge connection of each node includes: If the number of connections between the node in the weak connection and the nodes in the first community is greater than the number of connections between the node in the weak connection and the nodes in the second community, the node in the weak connection is divided into the first community; If the number of links between the nodes in the weak links and the nodes in the first community and the second community is equal, the average link centrality of the first community and the second community is calculated respectively, and the square with the smaller average link centrality is determined; The nodes in the weak links are divided into communities corresponding to the smaller average link centrality.

8. The community detection method in a network according to claim 1, characterized in that: The reallocation and adjustment of nodes in the community based on the edge connection of each node includes: Calculate and determine the redundancy of the nodes in the weak links in different communities, where the redundancy of the nodes in the community represents the number of links that are redundant compared to maintaining the hierarchical relationship of the first-order and second-order neighbors in the community; The nodes in the weak connection are divided into a target community with maximum redundancy.

9. A community detection device in a network, characterized in that: include: The contraction module is used to contract different edges in the target network to generate different new networks; A first determination module, configured to determine the edge centrality of the target network according to the difference between the target network and the new network; A second determination module is used to identify and determine weak edges and multiple communities in the target network according to the edge centrality, wherein the points in the weak edges are located at the edge of the community; An aggregation module, used to aggregate nodes in adjacent communities with the nodes in the weak links as starting points, and update and determine the edges of the communities based on the aggregation results; The adjustment module is used to reallocate and adjust the nodes in the community according to the edge connection status of each node, so as to obtain the detection result of the community in the target network.

10. An electronic device, characterized in that: include: one or more processors; a memory configured to store one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the community detection method within a network as described in any one of claims 1 to 8.

Images