Search method and system based on forbidden node perception

Abstract

The present disclosure proposes a search method and system based on forbidden node perception, including: obtaining a community network composed of multiple nodes and interactive relationships between nodes; assigning weights to the nodes in the community network and the interactive relationships between nodes according to the method of calculating the authority value of a web page; using a preset forbidden node sensitivity threshold to remove all nodes with weights less than the threshold from the community network; arranging the remaining nodes in descending order according to the size of the weights; adding nodes to the community network one by one in descending order of weights, and calculating the weighted derived degree corresponding to the community network at each addition; the community corresponding to the moment when the weighted derived degree is the smallest is the final result of the community search, which can help people find more accurate community results in the case of forbidden nodes.

Description

Technical Field

[0001] The disclosed embodiments relate to the technical field of computer network applications, and more particularly to a search method and system based on forbidden node perception. Background Art

[0002] In the fields of computer science, biology, sociology, etc., there are a large number of network structures formed by nodes and the connections between nodes. In network-related research, communities have received continuous attention. A community generally refers to a subgraph in which the connections between internal nodes are closer than the connections between nodes inside and outside the community. The community structure mined from the network helps people complete a series of tasks such as friend recommendation, criminal gang identification, and protein function prediction. The community search problem refers to finding communities containing one or more nodes in a given network.

[0003] Current community search methods mainly include methods based on specific topological structures and methods that comprehensively consider topological structures and node attributes. Typical topological structures include k-core and k-truss. The k-core structure requires that the degree of each node in the community is not less than k, while the k-truss requires that the number of triangles where each edge is located is not less than k-2. Most methods that comprehensively consider topological structures and node attributes comprehensively consider the similarity of node attributes. For example, the AGAR method first supplements the edges of the original graph based on the attribute similarity between nodes to construct a TA-graph. On this basis, the community search is performed based on the k-truss structure, and finally the community containing the given node is obtained. These methods focus on how to define the compact structure of the community and how to deal with node attributes, but do not consider people's constraints on the community other than containing necessary nodes.

[0004] In fact, existing community search methods can only find communities containing a given set of nodes, but do not allow people to impose more constraints on the nodes in the community. In actual application scenarios, people also hope that the searched community results do not contain certain given nodes, that is, forbidden nodes. In real life, forbidden nodes can correspond to people in a blacklist or products marked as disliked, etc., but existing methods lack support for this. To this end, an embodiment of the present invention aims to provide a search method based on forbidden node perception. Summary of the invention

[0005] In order to solve the above technical problems, the present disclosure proposes a search method based on forbidden node perception, including:

[0006] Obtain a community network consisting of multiple nodes and interaction relationships between nodes;

[0007] According to the method of calculating the authority value of web pages, weights are assigned to nodes and the interaction relationships between nodes in the community network;

[0008] Using a preset prohibited node sensitivity threshold, all nodes with a value less than the threshold are removed from the community network;

[0009] Arrange the remaining nodes in descending order according to their weights;

[0010] Add nodes to the community network one by one in descending order of weights, and calculate the weighted derived degree of the community network corresponding to each addition;

[0011] The community corresponding to the moment when the weighted derived degree is the smallest is the final result of the community search.

[0012] Furthermore, the weights assigned to the interaction between nodes in the community network according to the method for calculating the authority value of a web page represent the degree of remoteness of the interaction between nodes relative to the preset black nodes.

[0013] Furthermore, the method of calculating the authority value of a webpage to assign weights to nodes in a community network and the interaction relationships between nodes includes:

[0014] According to the number of nodes adjacent to each node, the average probability of interaction with each adjacent node is calculated;

[0015] According to the nodes that must be included in the community network, the webpage authority value is calculated to obtain the closeness of all nodes in the network to the nodes that must be included;

[0016] According to the black node users who are not allowed to appear in the community, the closeness of all nodes in the network to the black node is obtained by calculating the web page authority value;

[0017] Normalize the final remoteness of each node relative to the black node so that its value falls between 0 and 1. The value corresponding to the nodes that the community must include is set to 1 by default, and the value corresponding to the black node user is set to 0 by default. Calculate the final weight values ​​of the remaining nodes;

[0018] Calculate the interaction behavior between each node and the alienation degree of black nodes.

[0019] Furthermore, adding nodes to the community network one by one in descending order of weights and calculating the weighted derived degree corresponding to the community network each time the nodes are added includes:

[0020] All users of the social network are sorted in descending order according to their alienation from the black nodes, and then join the community one by one;

[0021] Whenever a node is added to the community, the temporary community network result is retained and the weighted derived degree corresponding to the community network result is calculated.

[0022] On the other hand, the present disclosure proposes a search system based on forbidden node perception, comprising:

[0023] An acquisition module is used to acquire a community network consisting of multiple nodes and interaction relationships between nodes;

[0024] A first calculation module, used to assign weights to nodes and interaction relationships between nodes in a community network according to a method for calculating web page authority values;

[0025] The first processing module uses a preset prohibited node sensitivity threshold to remove all nodes with a value less than the threshold from the community network;

[0026] The second processing module is used to arrange the remaining nodes in descending order according to the weight values;

[0027] The second calculation module adds nodes to the community network one by one in descending order of weights, and calculates the weighted derived degree corresponding to the community network each time the nodes are added;

[0028] The output module takes the community corresponding to the moment when the weighted export degree is the smallest as the final result of the community search.

[0029] Furthermore, the first calculation module assigns weights to the interaction relationships between nodes in the community network according to the method for calculating the authority value of a web page, indicating the degree of alienation between the interaction relationships between nodes relative to preset black nodes.

[0030] Furthermore, the first calculation module includes:

[0031] The first calculation submodule is used to calculate the average probability of each node interacting with each adjacent node according to the number of adjacent nodes of each node;

[0032] The first acquisition submodule is used to obtain the closeness of all nodes in the network to the nodes that must be included according to the nodes that must be included in the community network by calculating the web page authority value;

[0033] The second acquisition submodule is used to obtain the closeness of all nodes in the network to the black node by calculating the web page authority value according to the black node users who are not allowed to appear in the community;

[0034] The first processing submodule is used to normalize the value of the distance degree of each node relative to the black node so that its value falls between 0 and 1. The value corresponding to the node that the community must include is set to 1 by default, and the value corresponding to the black node user is set to 0 by default, and the final weight value of the remaining nodes is calculated;

[0035] The second processing submodule is used to calculate the interaction behavior between each node and the alienation degree of the black node.

[0036] Furthermore, the second calculation module includes:

[0037] The joining submodule is used to sort all users of the social network in descending order according to their alienation degree relative to the black nodes, and then join the community one by one;

[0038] The second calculation submodule is used to retain the temporary community network result and calculate the weighted derived degree corresponding to the community network result every time a node is added to the community.

[0039] Compared with the prior art, the present invention has the following advantages:

[0040] The present disclosure introduces new community search requirements into the community search problem, so that users can introduce forbidden nodes in the community search process to prevent unwanted people and things from appearing in the community results, such as blacklisted users, etc. The present disclosure can help people find more accurate community results in the presence of forbidden nodes. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] The accompanying drawings, which constitute a part of the specification, illustrate embodiments of the present disclosure and, together with the description, serve to explain the principles of the present disclosure.

[0042] The present disclosure may be more clearly understood from the following detailed description with reference to the accompanying drawings, in which:

[0043] Figure 1 It is a general flow chart of a community search method based on forbidden node perception in a social network of the present invention;

[0044] Figure 2 is a flow chart of the present invention using a personalized PageRank method to assign weights to users and relationships in a social network;

[0045] Figure 3 This is a flow chart of the present invention, according to the users that must be included in the community, obtaining the closeness of all users in the network to the users that must be included through a personalized PageRank method;

[0046] Figure 4 The present invention is a flow chart of obtaining the closeness of all users in the network to the blacklisted users through a personalized PageRank method according to the blacklisted users who are not allowed to appear in the community;

[0047] Figure 5 It is a flowchart for adding users to the community one by one in descending order of weights, and calculating the weighted derived degree corresponding to the community each time they join;

[0048] Figure 6 is a schematic diagram of weighting results for users in a social network using a personalized PageRank method in an embodiment of the present invention;

[0049] Figure 7 It is an experimental data diagram in an embodiment of the present invention. DETAILED DESCRIPTION

[0050] In order to make the purpose, technical solution and advantages of the embodiments of the present disclosure clearer, the technical solution in the embodiments of the present disclosure will be clearly and completely described below in conjunction with the drawings in the embodiments of the present disclosure. Obviously, the described embodiments 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 this field without creative work are within the scope of protection of the present disclosure.

[0051] The present disclosure proposes a search method based on forbidden node perception, comprising:

[0052] Obtain a community network consisting of multiple nodes and interaction relationships between nodes;

[0053] According to the method of calculating the authority value of web pages, weights are assigned to nodes and the interaction relationships between nodes in the community network;

[0054] Using a preset prohibited node sensitivity threshold, all nodes with a value less than the threshold are removed from the community network;

[0055] Arrange the remaining nodes in descending order according to their weights;

[0056] Add nodes to the community network one by one in descending order of weights, and calculate the weighted derived degree of the community network corresponding to each addition;

[0057] The community corresponding to the moment when the weighted derived degree is the smallest is the final result of the community search.

[0058] Specific reference Figure 1 ,A search method based on forbidden node perception in social network includes the following steps.

[0059] S1. Take several users who are active on a social platform (such as Twitter) within a period of time as nodes, and the interactions between users (such as likes, forwarding, etc.) as the relationships between nodes to construct a complex network. Usually, the network can be represented by an undirected graph G = (V, E), where V represents the set of n nodes in the graph, i.e., users, and E represents the set of m edges in the graph, i.e., the set of user interactions. In addition, according to the actual scenario needs, determine the users that must be included in the community to be searched and the blacklist users.

[0060] S2. A variation of PageRank, a common method for calculating the authority of web pages on the Internet, is used, namely the personalized PageRank method, to assign weights to users and interactions in social networks to indicate the degree of alienation between users and interactions and pre-specified blacklist users. For the specific process, see Figure 2 . Figure 6 This is a diagram of the weighted results of the social network obtained when users A and B are the users that must be included in the community and user C is a blacklisted user. The darker the color, the greater the weight of the node, that is, the closer it is to the nodes that must be included in the community and the more distant it is from the prohibited nodes.

[0061] S3. Use the preset forbidden node sensitivity threshold λ to remove all users with a value less than λ from the network, so that those blacklist users who do not want to appear in the community and users who are close to these blacklist users can be excluded from the network. The forbidden node sensitivity threshold λ is the preset minimum value of the alienation degree from the blacklist users, which can be adjusted according to actual needs. The value range is between 0 and 1. The larger the threshold, the smaller the scale of the community results that can be found. It is usually recommended to set it to 0.5.

[0062] S4. Arrange the remaining users in the social network in descending order according to their weights, that is, their degree of alienation from the blacklisted users, so as to clearly see their degree of closeness to the blacklisted users relative to the users that must be included.

[0063] S5. Add users to the community one by one in descending order of weights, and calculate the weighted derived degree of the community corresponding to each addition. The inventor believes that the structural tightness of the community can be reflected by the weighted derived degree, and defines a formula to calculate it. For the specific calculation process, see Figure 5 .

[0064] S6. The community corresponding to the moment when the weighted derived degree is the smallest is the final result of the community search. The smaller the weighted derived degree, the higher the structural compactness, so the community result at the moment when the weighted derived degree is the smallest in step S5 is taken as the final result of the community search. Figure 7 The method proposed in this invention (named FNACS) is compared with different benchmark methods (k-core method and k-truss method) on the DBLP dataset (paper collaboration network dataset) for 5 different community search results, which shows that the accuracy of community search (expressed by F-measure) of the method proposed in this invention is superior to that of the benchmark methods.

[0065] See also Figure 2, using a variation of PageRank, a common method for calculating web page authority on the Internet, namely the personalized PageRank method, to assign weights to users and relationships in social networks to represent the degree of alienation between users and relationships and pre-specified blacklist users, including the following steps:

[0066] S21. According to the number of users adjacent to each user, calculate the average probability of interaction between each user and each adjacent user, and integrate them into a transition probability matrix. The calculation method is M = A T D -1 , where M is the transition probability matrix, A is the adjacency matrix corresponding to the network graph, and D is a diagonal matrix whose diagonal elements are the number of users adjacent to each user.

[0067] S22. According to the users that must be included in the community, the degree of closeness of all users in the network to the users that must be included is obtained through the personalized PageRank method, with the vector p q presented in the form of.

[0068] S23. According to the blacklist users who are not allowed to appear in the community, the degree of closeness of all users in the network to the blacklist users is obtained through the personalized PageRank method, with the vector p f presented in the form of.

[0069] S24. Calculate the final alienation degree of each user relative to the blacklisted users. The calculation formula is as follows: P(v) represents the final alienation degree of user v relative to the blacklisted users, P q (v) represents the closeness of user v to the users that must be included in 22), P f (v) represents the closeness of user v to the blacklisted user in 23), and V is the set of all users:

[0070]

[0071] S25. Normalize the final value of the alienation degree of each user relative to the blacklisted users so that its value falls between 0 and 1. The values ​​corresponding to the users that must be included in the community are set to 1 by default, and the values ​​corresponding to the blacklisted users are set to 0 by default. The remaining users are calculated according to the following formula to calculate the final normalized weight value, denoted by W(v):

[0072]

[0073] S26. Calculate the distance between each user's interaction behavior and the blacklisted users. The specific calculation formula is as follows. The distance between each interaction behavior and the blacklisted users is represented by W(u,v), where u and v represent the two interacting users, and degree(u) and degree(v) represent the number of users interacting with user u and user v respectively:

[0074]

[0075] See also Figure 3 According to the users that must be included in the community, the personalized PageRank method is used to obtain the closeness of all users in the network to the users that must be included, with the vector p q The form of presentation includes the following steps:

[0076] S221. Set the probability of a certain user in the network directly interacting with the users that must be included in the community, denoted as the transmission probability α, which can generally be set to 0.15 according to experience. At the same time, set a unit vector v that represents the inclination towards the users that must be included in the community, each element of which corresponds to a user in the network, and only the elements corresponding to the users that must be included in the community are assigned the same non-zero positive value, and other users are assigned a value of 0 to represent the inclination of other users towards the users that must be included in the community.

[0077] S222. In the initial case, the closeness of all users to the community must be set to the same 1 / n, and the vector p q It is presented, called a personalized PageRank vector, and n is the number of users in the network.

[0078] S223. Initialize the personalized PageRank vector p q Through iterative calculation, we finally get a stable vector p q , the elements it contains are the closeness of each user to the users that the community must include. The specific calculation method is to convert the initial vector p q Substitute the formula p q =αv+(1-α)Mp q Get the new p on the right q Repeat the above operation several times until p q It is sufficient that it does not change or the degree of change is within the allowed range. In the formula, M is the transition probability matrix calculated in 21).

[0079] See also Figure 4 According to the blacklist users who are not allowed to appear in the community, the personalized PageRank method is used to obtain the closeness of all users in the network to the blacklist users, with the vector p fThe form of presentation includes the following steps:

[0080] S231. Set the probability of a certain user in the network directly interacting with a blacklist user, denoted as the transmission probability α, which can generally be set to 0.15 according to experience. At the same time, set a unit vector v representing the tendency towards blacklist users, each element of which corresponds to a user in the network, and only the elements corresponding to the blacklist users are assigned the same non-zero positive value, and other users are assigned a value of 0 to indicate the tendency of other users towards the blacklist users.

[0081] S232. In the initial case, the closeness of all users to the blacklisted users is set to the same 1 / n, and the vector p f It is presented, called a personalized PageRank vector, and n is the number of users in the network.

[0082] S233. Initialize the personalized PageRank vector p f Through iterative calculation, we finally get a stable vector p f , the elements it contains are the closeness of each user to the blacklisted users. The specific calculation method is to convert the initial vector p f Substitute the formula p f =αv+(1-α)Mp f Get the new p on the right f Repeat the above operation several times until p f It is sufficient that it does not change or the degree of change is within the allowed range. In the formula, M is the transition probability matrix calculated in 21).

[0083] See also Figure 5 , add users to the community one by one in descending order of weights, and calculate the weighted export degree corresponding to the community each time they join, including:

[0084] S51. After all the users of the social network are sorted in descending order according to their alienation from the blacklisted users, they are added to the community one by one.

[0085] S52. Whenever a user is added to the community, the temporary community result is retained and the weighted export degree corresponding to the community result is calculated, represented by WC(H), to measure the quality of the community result. The calculation formula of the weighted export degree is as follows:

[0086]

[0087] Among them, H represents the subgraph corresponding to the community, S represents the users in H, and W(u,v) is the degree of alienation between the interaction behavior mentioned in 26) and the blacklisted users.

[0088] Furthermore, the weights assigned to the interaction between nodes in the community network according to the method for calculating the authority value of a web page represent the degree of remoteness of the interaction between nodes relative to the preset black nodes.

[0089] Furthermore, the method of calculating the authority value of a webpage to assign weights to nodes in a community network and the interaction relationships between nodes includes:

[0090] According to the number of nodes adjacent to each node, the average probability of interaction with each adjacent node is calculated;

[0091] According to the nodes that must be included in the community network, the webpage authority value is calculated to obtain the closeness of all nodes in the network to the nodes that must be included;

[0092] According to the black node users who are not allowed to appear in the community, the closeness of all nodes in the network to the black node is obtained by calculating the web page authority value;

[0093] Normalize the final remoteness of each node relative to the black node so that its value falls between 0 and 1. The value corresponding to the nodes that the community must include is set to 1 by default, and the value corresponding to the black node user is set to 0 by default. Calculate the final weight values ​​of the remaining nodes;

[0094] Calculate the interaction behavior between each node and the alienation degree of black nodes.

[0095] Furthermore, adding nodes to the community network one by one in descending order of weights and calculating the weighted derived degree corresponding to the community network each time the nodes are added includes:

[0096] All users of the social network are sorted in descending order according to their alienation from the black nodes, and then join the community one by one;

[0097] Whenever a node is added to the community, the temporary community network result is retained and the weighted derived degree corresponding to the community network result is calculated.

[0098] On the other hand, the present disclosure proposes a search system based on forbidden node perception, comprising:

[0099] An acquisition module is used to acquire a community network consisting of multiple nodes and interaction relationships between nodes;

[0100] A first calculation module, used to assign weights to nodes and interaction relationships between nodes in a community network according to a method for calculating web page authority values;

[0101] The first processing module uses a preset prohibited node sensitivity threshold to remove all nodes with a value less than the threshold from the community network;

[0102] The second processing module is used to arrange the remaining nodes in descending order according to the weight values;

[0103] The second calculation module adds nodes to the community network one by one in descending order of weights, and calculates the weighted derived degree corresponding to the community network each time the nodes are added;

[0104] The output module takes the community corresponding to the moment when the weighted export degree is the smallest as the final result of the community search.

[0105] Furthermore, the first calculation module assigns weights to the interaction relationships between nodes in the community network according to the method for calculating the authority value of a web page, indicating the degree of alienation between the interaction relationships between nodes relative to preset black nodes.

[0106] Furthermore, the first calculation module includes:

[0107] The first calculation submodule is used to calculate the average probability of each node interacting with each adjacent node according to the number of adjacent nodes of each node;

[0108] The first acquisition submodule is used to obtain the closeness of all nodes in the network to the nodes that must be included according to the nodes that must be included in the community network by calculating the web page authority value;

[0109] The second acquisition submodule is used to obtain the closeness of all nodes in the network to the black node by calculating the web page authority value according to the black node users who are not allowed to appear in the community;

[0110] The first processing submodule is used to normalize the value of the distance degree of each node relative to the black node so that its value falls between 0 and 1. The value corresponding to the node that the community must include is set to 1 by default, and the value corresponding to the black node user is set to 0 by default, and the final weight value of the remaining nodes is calculated;

[0111] The second processing submodule is used to calculate the interaction behavior between each node and the alienation degree of the black node.

[0112] Furthermore, the second calculation module includes:

[0113] The joining submodule is used to sort all users of the social network in descending order according to their alienation degree relative to the black nodes, and then join the community one by one;

[0114] The second calculation submodule is used to retain the temporary community network result and calculate the weighted derived degree corresponding to the community network result every time a node is added to the community.

[0115] The present disclosure introduces new community search requirements into the community search problem, so that users can introduce forbidden nodes in the community search process to prevent unwanted people and things from appearing in the community results, such as blacklisted users. Figure 7 As shown, compared with the existing community search method, the present invention can help people find more accurate community results in the case of prohibited nodes.

[0116] For those skilled in the art, it is obvious that the embodiments of the present disclosure are not limited to the details of the above exemplary embodiments, and the embodiments of the present disclosure can be implemented in other specific forms without departing from the spirit or basic features of the embodiments of the present disclosure. Therefore, no matter from which point of view, the embodiments should be regarded as exemplary and non-restrictive, and the scope of the embodiments of the present disclosure is limited by the attached claims rather than the above description, so it is intended to include all changes that fall within the meaning and scope of the equivalent elements of the claims in the embodiments of the present disclosure. Any figure mark in the claims should not be regarded as limiting the claims involved. In addition, it is obvious that the word "including" does not exclude other units or steps, and the singular does not exclude the plural. Multiple units, modules or devices stated in the system, device or terminal claims can also be implemented by the same unit, module or device through software or hardware. The words first, second, etc. are used to indicate names, and do not indicate any specific order.

Claims

1. A search method based on forbidden node perception, characterized in that: include: Obtain a community network consisting of multiple nodes and interaction relationships between nodes; The weights assigned to the nodes and the interaction relationships between the nodes in the community network according to the method for calculating the webpage authority value represent the degree of alienation of the interaction relationships between the nodes relative to the preset black nodes. The method of calculating the authority value of a web page to assign weights to nodes and the interaction relationships between nodes in a community network includes: According to the number of nodes adjacent to each node, the average probability of interaction with each adjacent node is calculated; According to the nodes that must be included in the community network, the webpage authority value is calculated to obtain the closeness of all nodes in the network to the nodes that must be included; According to the black node users who are not allowed to appear in the community, the closeness of all nodes in the network to the black node is obtained by calculating the web page authority value; Normalize the final remoteness of each node relative to the black node so that its value falls between 0 and 1. The value corresponding to the nodes that the community must include is set to 1 by default, and the value corresponding to the black node user is set to 0 by default. Calculate the final weight values ​​of the remaining nodes; Calculate the interaction behavior between each node and the alienation degree of black nodes; Using a preset prohibited node sensitivity threshold, all nodes with a value less than the threshold are removed from the community network; Arrange the remaining nodes in descending order according to their weights; Adding nodes to the community network one by one in descending order of weights, and calculating the weighted derived degree corresponding to the community network each time the nodes are added; adding nodes to the community network one by one in descending order of weights, and calculating the weighted derived degree corresponding to the community network each time the nodes are added includes: All users of the social network are sorted in descending order according to their alienation from the black nodes, and then join the community one by one; Whenever a node is added to the community, the temporary community network result is retained and the weighted derived degree corresponding to the community network result is calculated; The community corresponding to the moment when the weighted derived degree is the smallest is the final result of the community search.

2. A search system based on forbidden node perception, characterized in that: include: An acquisition module is used to acquire a community network consisting of multiple nodes and interaction relationships between nodes; A first calculation module, used to assign weights to nodes and interaction relationships between nodes in a community network according to a method for calculating web page authority values; The first calculation module assigns weights to the interaction relationships between nodes in the community network according to the method for calculating the authority value of a web page, indicating the degree of alienation between the interaction relationships between nodes relative to the preset black nodes; The first calculation module includes: The first calculation submodule is used to calculate the average probability of each node interacting with each adjacent node according to the number of adjacent nodes of each node; The first acquisition submodule is used to obtain the closeness of all nodes in the network to the nodes that must be included according to the nodes that must be included in the community network by calculating the web page authority value; The second acquisition submodule is used to obtain the closeness of all nodes in the network to the black node by calculating the web page authority value according to the black node users who are not allowed to appear in the community; The first processing submodule is used to normalize the value of the distance degree of each node relative to the black node so that its value falls between 0 and 1. The value corresponding to the node that the community must include is set to 1 by default, and the value corresponding to the black node user is set to 0 by default, and the final weight value of the remaining nodes is calculated; The second processing submodule is used to calculate the interaction behavior between each node and the alienation degree of the black node; The first processing module uses a preset prohibited node sensitivity threshold to remove all nodes with a value less than the threshold from the community network; The second processing module is used to arrange the remaining nodes in descending order according to the weight values; The second calculation module adds nodes to the community network one by one in descending order of weights, and calculates the weighted derived degree corresponding to the community network each time the nodes are added; the second calculation module includes: The joining submodule is used to sort all users of the social network in descending order according to their alienation degree relative to the black nodes, and then join the community one by one; The second calculation submodule is used for retaining the temporary community network result and calculating the weighted derived degree corresponding to the community network result whenever a node is added to the community; The output module takes the community corresponding to the moment when the weighted export degree is the smallest as the final result of the community search.

Images