Method and related device for hiding network community based on user attribute features

Abstract

The application belongs to the field of network community, and discloses a network community hiding method based on user attribute characteristics and related devices, which comprises the following steps: obtaining the edge addition selection probability of each community pair according to the separation degree, attribute similarity and the number of nodes of each community in the community pair; obtaining the edge deletion selection probability of each community according to the cohesion degree, attribute uniqueness and the number of node pairs; randomly selecting a community pair as a target community pair according to the edge addition selection probability of each community pair, and adding an edge to the non-edge with the minimum product of the sum of node degrees and (1-attribute similarity) in the target community pair; and randomly selecting a community as a target community according to the edge deletion selection probability of each community, and deleting an edge with the maximum product of the sum of node degrees and (1-attribute similarity) in the target community. The method breaks through the limitation of a specific target community mode, effectively overcomes the drawbacks that existing algorithms generally ignore node attribute information, and improves the community hiding performance.

Description

Technical Field

[0001] This invention belongs to the field of online communities and relates to a method and related apparatus for hiding online communities based on user attribute characteristics. Background Technology

[0002] In real-world graph network communities, node attributes contain rich user information, such as gender, age, and interests. These attributes are crucial for accurately identifying and hiding communities. However, most existing technologies focus on perturbing the community structure, neglecting the mining and utilization of node attribute information. This makes current methods unable to fully consider users' personalized characteristics when hiding communities, potentially leading to poor hiding results. Once community detection algorithms utilize this neglected attribute information, users are easily detected, leading to privacy breaches and seriously compromising user information security. Summary of the Invention

[0003] The purpose of this invention is to overcome the shortcomings of the prior art and provide a method and related apparatus for hiding online communities based on user attribute characteristics.

[0004] To achieve the above objectives, the present invention employs the following technical solution: In a first aspect, the present invention provides a method for hiding network communities based on user attribute features, comprising: acquiring the graph structure, community information, and node attribute information of the network community; based on the graph structure, community information, and node attribute information of the network community, for each community pair within the network community, acquiring and obtaining the edge addition selection probability of each community pair based on the separation degree, attribute similarity, and the number of nodes in each community within the community pair; for each community within the network community, acquiring and obtaining the edge deletion selection probability of each community based on the cohesion, attribute uniqueness, and number of node pairs in each community; randomly selecting a community pair as a target community pair based on the edge addition selection probability of each community pair, and selecting the non-edge with the smallest product of the sum of node degrees and (1-attribute similarity) in the target community pair for edge addition; randomly selecting a community as a target community based on the edge deletion selection probability of each community, and selecting the edge with the largest product of the sum of node degrees and (1-attribute similarity) in the target community for edge deletion.

[0005] Optionally, obtaining the edge addition selection probability of each community pair based on the separation degree, attribute similarity, and number of nodes in each community within the community pair includes: obtaining the attribute similarity of each community pair by the following formula: taking the cosine similarity between the average attribute vectors of the two communities within the community pair as the attribute similarity of the community pair.

[0006] Optionally, obtaining the edge addition selection probability of each community pair based on the separation degree, attribute similarity, and the number of nodes in each community within the community pair includes: obtaining the edge addition selection probability of each community pair using the following formula: P J（ij） =|X i ||X j |×S(X i X j )×

[0007] Among them, P J（ij） For the community to (X) i X j Add selection probability to the edge of X; i X is the community numbered i in the online community; j For community number j in the online community; |X i |For X i The number of nodes; |X j |For X j The number of nodes; S(X) i X j ) for the community to (X) i X j The resolution of ) For the community to (X) i X j The similarity of attributes.

[0008] Optionally, obtaining the edge deletion selection probability of each community based on its cohesion, attribute uniqueness, and number of node pairs includes obtaining the attribute uniqueness of each community in the following way: taking the cosine similarity between the average attribute vector of all communities in the network community and the average attribute vector of the current community as the attribute uniqueness of the current community.

[0009] Optionally, obtaining the edge deletion selection probability of each community based on its cohesion, attribute uniqueness, and number of node pairs includes: obtaining the edge deletion selection probability of each community using the following formula: P S（i） =C(X i ) ×|X i ||X i -1| Among them, P S（i） For X i The edge deletion selection probability; X i C(X) is the community numbered i in the online community. i ) is X i The degree of cohesion; For the uniqueness of community attributes; |Xi ||X i -1| is twice the number of node pairs in the community.

[0010] Optionally, when adding edges to the non-edges that minimize the product of the sum of node degrees and (1-attribute similarity) of the target community, if there are at least two non-edges that minimize the product of the sum of node degrees and (1-attribute similarity) of the target community: select the non-edge that minimizes the increase in the external degree score of the network community from the non-edges that minimize the product of the sum of node degrees and (1-attribute similarity) of the target community.

[0011] Optionally, when selecting the edge with the largest product of the sum of node degrees and (1-attribute similarity) in the target community for edge deletion, if there are at least two edges with the largest product of the sum of node degrees and (1-attribute similarity) in the target community: select the edge that maximizes the increase in the external degree score of the network community among the edges with the largest product of the sum of node degrees and (1-attribute similarity) in the target community for edge deletion.

[0012] In a second aspect, the present invention provides a network community hiding system based on user attribute features, comprising: a data acquisition module for acquiring the graph structure, community information, and node attribute information of the network community; a budget allocation module for, based on the graph structure, community information, and node attribute information of the network community, acquiring and obtaining the edge addition selection probability for each community pair within the network community according to the separation degree, attribute similarity, and number of nodes in each community within the community pair; acquiring and obtaining the edge deletion selection probability for each community within the network community according to the cohesion, attribute uniqueness, and number of node pairs in each community; an edge addition module for, randomly selecting a community pair as a target community pair according to the edge addition selection probability of each community pair, and selecting the non-edge with the smallest product of the sum of node degrees and (1-attribute similarity) in the target community pair for edge addition; and an edge deletion module for, randomly selecting a community as a target community according to the edge deletion selection probability of each community, and selecting the edge with the largest product of the sum of node degrees and (1-attribute similarity) in the target community for edge deletion.

[0013] In a third aspect, the present invention provides a computer device including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements the steps of the above-described network community hiding method based on user attribute features.

[0014] In a fourth aspect, the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the above-described network community hiding method based on user attribute features.

[0015] Compared with the prior art, the present invention has the following beneficial effects: This invention presents a network community hiding method based on user attribute features. It innovatively integrates node attributes with community structure information. By leveraging the separation degree of each community pair, attribute similarity, and the number of nodes in each community within a community pair, the probability of adding edges to each community pair is determined. Simultaneously, by utilizing the cohesion, attribute uniqueness, and number of node pairs within each community, the probability of deleting edges from each community is determined. Furthermore, during edge perturbation, introducing edges between highly similar node pairs allows nodes to more easily migrate to other communities; simultaneously, deleting edges within a community further strengthens the separation between nodes, making it easier for nodes to detach from their original communities. These processes significantly improve the community hiding effect with a smaller perturbation budget, efficiently evading community detection algorithms and ensuring the security of user attribute information. Unlike traditional methods, this invention breaks through the limitations of specific target community patterns, flexibly addressing a wider range of community hiding needs and greatly enhancing the flexibility of hiding arbitrary target communities. By fully mining the structural and node attribute information within the community, this method effectively overcomes the shortcomings of existing algorithms that generally ignore node attribute information. This allows for better handling of complex community relationships, significantly improves community hiding performance, effectively prevents users from being identified by detection algorithms, greatly reduces the risk of user privacy leaks, and truly protects the privacy and security of online community users. Attached Figure Description

[0016] Figure 1 This is a flowchart of a network community hiding method based on user attribute features according to an embodiment of the present invention.

[0017] Figure 2 This is a block diagram of the network community hiding system based on user attribute features according to an embodiment of the present invention. Detailed Implementation

[0018] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0019] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0020] The present invention will now be described in further detail with reference to the accompanying drawings: See Figure 1 In one embodiment of the present invention, a network community hiding method based on user attribute features is provided, which innovatively integrates node attributes with community structure information to achieve efficient community hiding.

[0021] Specifically, the network community hiding method based on user attribute features of this invention includes the following steps: S1: Obtain the graph structure, community information, and node attribute information of the network community; S2: Based on the graph structure, community information, and node attribute information of the network community, for each community pair within the network community, obtain and based on the separation degree, attribute similarity, and number of nodes in each community within the community pair, obtain the edge addition selection probability of each community pair; for each community within the network community, obtain and based on the cohesion, attribute uniqueness, and number of node pairs of each community, obtain the edge deletion selection probability of each community. S3: Randomly select community pairs as target community pairs based on the edge addition selection probability of each community pair, and select the non-edges with the smallest product of the sum of node degrees and (1-attribute similarity) in the target community pair for edge addition; S4: Randomly select a community as the target community based on the edge deletion selection probability of each community, and select the edge with the largest sum of node degree and (1-attribute similarity) product in the target community for edge deletion.

[0022] This invention presents a network community hiding method based on user attribute features. It innovatively integrates node attributes with community structure information. By leveraging the separation degree of each community pair, attribute similarity, and the number of nodes in each community within a community pair, the probability of adding edges to each community pair is determined. Simultaneously, by utilizing the cohesion, attribute uniqueness, and number of node pairs within each community, the probability of deleting edges from each community is determined. Furthermore, during edge perturbation, introducing edges between highly similar node pairs allows nodes to more easily migrate to other communities; simultaneously, deleting edges within a community further strengthens the separation between nodes, making it easier for nodes to detach from their original communities. These processes significantly improve the community hiding effect with a relatively small perturbation budget, efficiently evading community detection algorithms and ensuring the security of user attribute information.

[0023] Explaining this further, current methods for hiding online communities, such as modular approaches, prioritize adding edges between the two communities that maximize the increase in total community volume. This strategy leads to uneven distribution of perturbation resources, with some target communities receiving excessive attention while others are easily overlooked, increasing the risk of exposure. Furthermore, existing methods employ only a simple random strategy when selecting which edges to add or remove, lacking an effective differentiation mechanism. This makes some users more vulnerable to detection algorithms, further exacerbating the possibility of privacy breaches and failing to guarantee user privacy and security.

[0024] Furthermore, modularization methods have limitations in measuring community structure, sometimes failing to accurately reflect the true state of the community. In some networks with unclear structures, modularity values ​​may be overestimated, leading to poor performance of modularization-based deception strategies. Simultaneously, existing methods struggle to achieve ideal performance and efficiency when handling complex graph networks and diverse community structures. They often find it difficult to strike a good balance between hiding effectiveness and computational cost, either incurring high computational costs to achieve high hiding effectiveness or suffering from suboptimal hiding effectiveness due to computational resource constraints. This results in an inability to reliably protect user privacy and effectively address the excessive exploitation of user privacy by community detection algorithms.

[0025] This invention provides a novel and more efficient community hiding strategy. Unlike traditional methods, this method breaks through the limitations of specific target community patterns and can flexibly address a wider range of community hiding needs, greatly improving the flexibility of hiding arbitrary target communities. By fully mining the structural information and node attribute information within the community, it effectively overcomes the shortcomings of existing algorithms that generally ignore node attribute information. This allows for better handling of complex community relationships, significantly improving community hiding performance, and thus effectively preventing users from being identified by detection algorithms, greatly reducing the risk of user privacy leakage, and effectively protecting the privacy and security of online community users.

[0026] Explained, when adding and deleting simultaneously, this invention, based on preset values ​​for the number of additions and deletions, completes the hiding of the network community based on user attribute characteristics when the number of additions and deletions reaches the preset values.

[0027] In one possible implementation, the graph structure, community information, and node attribute information of the network community are represented by a network such as G=(V,E) and partitions P={ The network is represented by the structure}, where V represents the set of nodes and E represents the set of edges in the network. These edges connect the nodes and form the basic structure of the network community. This setting lays the foundation for subsequent operations, ensuring a clear understanding of the entire network community environment.

[0028] In one possible implementation, obtaining the edge addition selection probability of each community pair based on the separation degree, attribute similarity, and number of nodes in each community within the community pair includes: obtaining the attribute similarity of each community pair by the following formula: taking the cosine similarity between the average attribute vectors of the two communities within the community pair as the attribute similarity of the community pair.

[0029] For interpretation, the range of cosine similarity is [-1, 1], and the larger the value, the higher the similarity.

[0030] Optionally, obtaining the edge addition selection probability of each community pair based on the separation degree, attribute similarity, and the number of nodes in each community within the community pair includes: obtaining the edge addition selection probability of each community pair using the following formula: P J（ij） =|X i ||X j |×S(X i X j )×

[0031] Among them, P J（ij） For the community to (X) i X j Add selection probability to the edge of X; i X is the community numbered i in the online community; j For community number j in the online community; |X i |For X i The number of nodes; |X j |For X j The number of nodes; S(X) i X j ) for the community to (X) i X j The resolution of ) For the community to (X) i Xj The similarity of attributes.

[0032] Explained, the edge addition selection probability for each community pair is essentially a rational allocation of the edge addition budget. Adding cross-community edges increases the probability of merging two communities. To fairly and efficiently allocate the edge addition budget, it is necessary to consider the probability of merging two communities after adding a new edge and the benefits of the merger. Separation degree is... Measuring inter-community ( The sparsity of connections indicates the degree of separation; the lower the separation, the sparser the connections, and the more difficult it is to merge. Therefore, it can be used as a measure of merging. and The difficulty index. If the two communities merge, Each node will receive | | New friends in the same community; in addition, the similarity of attributes between the two communities will affect the formation of new connections after the merger.

[0033] In this implementation, the cosine similarity between the average attribute vectors of two communities within a community pair is used as the attribute similarity of the community pair. The higher the attribute similarity, the easier it is to form a connection after merging. Therefore | || | This reflects the benefits of the merger. Taking into account both costs and benefits, a community pair is randomly selected in each round. , ), its probability is related to |X i ||X j |×S(X i X j )× Proportional (excluding non-edges without endpoints outside the target community set). In this way, the edge addition budget can be more reasonably allocated to each community pair, avoiding poor hiding effects in some communities due to uneven resource allocation, thereby reducing the risk of user privacy leakage.

[0034] Based on the above design, when performing edge addition operations, the possibility of community merging (measured by separation degree) and the benefits brought by merging (factors such as attribute similarity) can be comprehensively considered. This allows for a reasonable allocation of the budget and avoids excessive concentration of perturbation resources in a few communities. The core purpose of adding edges is to make it easier for target communities to merge with other communities. The above method accurately measures the possibility of two communities merging after adding edges and clearly defines the benefits brought by merging, such as the number of new neighbors added to community members after merging, and the lower the difficulty of forming new connections after merging if the two communities have higher attribute similarity. In this way, the problem of uneven distribution of perturbation resources in existing technologies is effectively solved, ensuring that each target community can obtain reasonable perturbation resources, thereby comprehensively improving the overall concealment effect and minimizing the possibility of privacy leaks caused by unreasonable resource allocation for individual network community users.

[0035] In one possible implementation, obtaining the edge deletion selection probability of each community based on its cohesion, attribute uniqueness, and number of node pairs includes: obtaining the attribute uniqueness of each community by using the cosine similarity between the average attribute vector of all communities in the network community and the average attribute vector of the current community as the attribute uniqueness of the current community.

[0036] Optionally, obtaining the edge deletion selection probability of each community based on its cohesion, attribute uniqueness, and number of node pairs includes: obtaining the edge deletion selection probability of each community using the following formula: P S（i） =C(X i ) ×|X i ||X i -1| Among them, P S（i） For X i The edge deletion selection probability; X i C(X) is the community numbered i in the online community. i ) is X i The degree of cohesion; For the uniqueness of community attributes; |X i ||X i -1| is twice the number of node pairs in the community.

[0037] Explanatoryly, the edge deletion selection probability of each community is essentially a rational allocation of the edge deletion budget; deleting edges within a community increases the likelihood of community splitting. Cohesion is... Measuring the difficulty of splitting The higher the value, the more pronounced the community structure, requiring a larger perturbation budget to effectively conceal it. Regarding benefit assessment, it is believed that if the community... After the split, its nodes can be better hidden in other communities, even if they gain benefits.

[0038] Specifically, the cosine similarity between the average attribute vector of all communities within the network and the average attribute vector of the current community is used as the attribute uniqueness of the current community. The number of potential node pairs within the community is then used to quantify the reward. A higher reward value indicates a stronger community. The weaker the structural uniqueness of a node, the easier it is for its nodes to hide within other communities. Therefore, a community is randomly selected each round. Its probability is related to C(X). i ) ×|X i ||X i -1 is directly proportional. This edge-deletion budget allocation strategy can precisely disturb the community, making it more prone to splitting, thereby disrupting the original community structure and reducing the likelihood of community detection algorithms identifying user privacy information.

[0039] Explanatoryly, when selecting communities for edge deletion, both the difficulty of splitting (measured by cohesion) and potential benefits (involving attribute uniqueness and the number of node pairs) are comprehensively considered. Through a comprehensive analysis of these factors, communities for edge deletion are rationally selected to achieve the goal of making the community easier to split. In this process, the costs and benefits of edge deletion are clearly defined. The cost reflects the stability of the community's own structure, used to measure the difficulty of splitting the community due to edge deletion; the benefit reflects whether internal nodes can be better hidden in other communities after the split. This effectively solves the "Matthew effect" that may occur in budget allocation, avoiding resource imbalance. By rationally selecting communities for deletion, while reducing community stability, the possibility of node dispersion and hiding is increased, thereby significantly reducing the probability of users being detected by the algorithm, further reducing the risk of user privacy leakage, and providing more reliable protection for user privacy.

[0040] In one possible implementation, when adding edges to the non-edges that minimize the product of the sum of the node degrees of the target community and (1-attribute similarity), if there are at least two non-edges that minimize the product of the sum of the node degrees of the target community and (1-attribute similarity): select the non-edge that minimizes the increase in the external degree score of the network community from the non-edges that minimize the product of the sum of the node degrees of the target community and (1-attribute similarity).

[0041] Explanatoryly, when adding edges, an edge addition strategy based on node degree and node attribute similarity is adopted. Let the eigenvalues ​​of the Laplacian matrix L of the network graph G be... From a graph cut perspective, to make the community structure of the subgraph G=(X∪Y,E) composed of two communities X and Y more difficult to detect, its algebraic connectivity should be increased, i.e., maximized. ( ),in, This is the perturbation graph after adding edges. However, it generally maximizes algebraic connectivity. It is an NP-hard problem (a nondeterministic polynomial-time problem). According to perturbation theory, if Isolated, for any non-edge (i, j) ∈ The additional first-order approximation is given, where, Let F(x) be the Fiedler vector of graph G. However, methods based on perturbation theory are computationally limited in practical scenarios. Therefore, by defining F(x) = (x - (1 / n))², we can... The maximization problem is transformed into:

[0042] in, Let be the degree of the k-th node, and m be the number of original edges in graph G. Add an edge between node pairs (i,j), where F is a defined function. , It is a non-edge set. For community X and Y.

[0043] Further derivation shows that this problem is equivalent to:

[0044] in, , It is a non-edge set.

[0045] Next, node attribute information is incorporated into the edge addition process. Specifically, when constructing the edge set, cosine similarity is used to accurately quantify the similarity between node attribute vectors. A higher cosine similarity indicates that, after adding an edge, the corresponding node pair is more likely to be hidden within the same community. Therefore, edges are added to non-edges in the target community pair whose sum of node degrees multiplied by (1 - attribute similarity) has the smallest value. This is because adding edges between nodes with small attribute differences is more likely to disrupt the original community structure, thereby enhancing the community hiding effect and effectively preventing user privacy from being obtained by detection algorithms.

[0046] Explanatoryly, when selecting specific edges to add after choosing community pairs, a comprehensive consideration is given to the connectivity (sum of node degrees) and attribute similarity of the nodes. Edges are preferentially added to node pairs with low connectivity and high attribute similarity because adding edges between such pairs is more likely to disrupt the original community structure, blurring the boundaries between communities and thus achieving the goal of hiding the communities.

[0047] Among these, when there are at least two non-edges that minimize the product of the sum of node degrees and (1 - attribute similarity) in the target community, the non-edges that minimize the increase in the external degree score (ODF) of the network community are selected, where the ODF of node u is defined as: Where u and v represent nodes, and S is a subset of nodes. Let be the degree of node u.

[0048] In one possible implementation, when selecting the edge with the largest sum of node degrees and (1-attribute similarity) product in the target community for edge deletion, if there are at least two edges with the largest sum of node degrees and (1-attribute similarity) product in the target community: select the edge that maximizes the increase in the external degree score of the network community among the edges with the largest sum of node degrees and (1-attribute similarity) product in the target community for edge deletion.

[0049] Explanatoryly, when performing edge deletion, a strategy based on edge perturbation of node degree and node attribute similarity is adopted. For edge deletion within a single community X, the goal is to reduce connectivity within the community. Consider the complement graph of G. Deleting an edge in G is equivalent to... Add an edge to make The nodes in G are more tightly connected, making community X in G more prone to splitting. Therefore, maximizing algebraic connectivity ( This is equivalent to minimizing the maximum eigenvalue of G. ( ),in, This is the perturbation graph after removing edges. It can also be approximated as:

[0050] Where E(X) is the edge set within the community X, and G is the original network structure. To delete the edge between (i,j), where m is the number of edges in G. Let f be the k-th eigenvalue of the Laplacian matrix, and F be a defined function. .

[0051] Further equivalent to:

[0052] Where E(X) is the edge set within the community X. It is the sum of the node degrees of the node pair (i, j).

[0053] This involves selecting the edge that maximizes the increase in the external degree score of the network community from the edges whose sum of node degrees multiplied by (1 - attribute similarity) is the largest. This edge is then deleted because deleting such an edge will significantly affect the connectivity of the community and greatly increase the probability of community splitting, thereby achieving the expected goal of community hiding and effectively preventing the leakage of user privacy information.

[0054] Specifically, when there are at least two edges in the target community whose sum of node degrees multiplied by (1 - attribute similarity) maximizes the product, the edge that maximizes the increase in ODF is selected from these edges. The edge that maximizes the increase in ODF is defined as... , Let be the out-degree fraction of node u. Let be the out-degree score of node v. , Let u and v be the degrees of nodes.

[0055] Explaining this, in transforming the spectral optimization problem into an edge volume optimization problem, we fully utilize node attribute information and combine it with node degree information for edge selection. When adding edges, we select non-edges with the smallest sum of node degrees and high attribute similarity; when deleting edges, we select edges with the largest sum of node degrees and low attribute similarity. For example, heuristic algorithms can be used to further improve operational efficiency, such as by sorting to reduce the search space and dynamically updating data to optimize the algorithm execution process.

[0056] In summary, the edge-adding and edge-deleting strategy of this invention has two main advantages: First, the edge selection strategy based on intuitive statistical information such as node degree and node attribute information can more accurately perturb the community structure, making the community structure more ambiguous in the face of detection algorithms, effectively improving the deception effect, greatly reducing the risk of users being detected by the algorithm, and thus reducing the possibility of privacy leakage. Second, the efficient algorithm design significantly reduces computational costs and greatly improves the algorithm's execution efficiency in the network. For example, when processing social network data, it can quickly complete the hiding of target communities while maintaining low computational resource consumption. This successfully overcomes the performance and efficiency deficiencies of existing methods in network applications, providing strong support for user privacy protection in large-scale network environments.

[0057] The following are embodiments of the apparatus of the present invention, which can be used to execute embodiments of the method of the present invention. For details not disclosed in the apparatus embodiments, please refer to the embodiments of the method of the present invention.

[0058] See Figure 2In another embodiment of the present invention, a network community hiding system based on user attribute features is provided, which can be used to implement the above-mentioned network community hiding method based on user attribute features. Specifically, the network community hiding system based on user attribute features includes a data acquisition module, a budget allocation module, an add-while-delete module, and an delete-while-add module.

[0059] The data acquisition module is used to acquire the graph structure, community information, and node attribute information of the network community. The budget allocation module is used to acquire, based on the graph structure, community information, and node attribute information of the network community, for each community pair within the network community, and obtain the edge addition selection probability of each community pair according to the separation degree, attribute similarity, and number of nodes in each community within the community pair. For each community within the network community, the module acquires and obtains the edge deletion selection probability of each community according to the cohesion, attribute uniqueness, and number of node pairs in each community. The edge addition module is used to randomly select community pairs as target community pairs according to the edge addition selection probability of each community pair, and select the non-edges with the smallest product of the sum of node degrees and (1-attribute similarity) in the target community pair for edge addition. The edge deletion module is used to randomly select communities as target communities according to the edge deletion selection probability of each community, and select the edge with the largest product of the sum of node degrees and (1-attribute similarity) in the target community for edge deletion.

[0060] All relevant content of each step involved in the aforementioned embodiments of the network community hiding method based on user attribute features can be referenced to the functional description of the corresponding functional module of the network community hiding system based on user attribute features in the embodiments of the present invention, and will not be repeated here.

[0061] The module division in this embodiment of the invention is illustrative and represents only one logical functional division. In actual implementation, other division methods may be used. Furthermore, the functional modules in the various embodiments of the invention can be integrated into a single processor, exist as separate physical entities, or be integrated into a single module. The integrated modules described above can be implemented in hardware or as software functional modules.

[0062] In another embodiment of the present invention, a computer device is provided, comprising a processor and a memory. The memory stores a computer program, which includes program instructions. The processor executes the program instructions stored in the computer storage medium. The processor may be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. It is the computing and control core of the terminal, suitable for implementing one or more instructions, specifically suitable for loading and executing one or more instructions in the computer storage medium to achieve a corresponding method flow or corresponding function. The processor described in this embodiment of the present invention can be used for the operation of a network community hiding method based on user attribute characteristics.

[0063] In another embodiment of the present invention, a storage medium is provided, specifically a computer-readable storage medium (Memory), which is a memory device in a computer device used to store programs and data. It is understood that the computer-readable storage medium here can include both the built-in storage medium in the computer device and extended storage media supported by the computer device. The computer-readable storage medium provides storage space that stores the terminal's operating system. Furthermore, the storage space also stores one or more instructions suitable for loading and execution by a processor. These instructions can be one or more computer programs (including program code). It should be noted that the computer-readable storage medium here can be high-speed RAM or non-volatile memory, such as at least one disk storage device. The processor can load and execute one or more instructions stored in the computer-readable storage medium to implement the corresponding steps of the network community hiding method based on user attribute characteristics in the above embodiments.

[0064] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0065] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0066] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0067] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0068] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A method for hiding network community attributes based on user attributes, characterized in that, include: Obtain the graph structure, community information, and node attribute information of the network community; Based on the graph structure, community information, and node attribute information of the network community, for each community pair within the network community, the edge addition selection probability of each community pair is obtained and based on the separation degree, attribute similarity, and the number of nodes in each community within the community pair. For each community within the network community, obtain and based on the cohesion, attribute uniqueness, and number of node pairs of each community, obtain the edge deletion selection probability of each community; Based on the edge addition selection probability of each community pair, community pairs are randomly selected as target community pairs. The non-edges with the smallest product of the sum of node degrees and (1 - attribute similarity) in the target community pairs are selected for edge addition. Based on the edge deletion selection probability of each community, a community is randomly selected as the target community. The edge with the largest sum of node degree and (1-attribute similarity) product in the target community is selected for edge deletion.

2. The network community hiding method based on user attribute features according to claim 1, characterized in that, The process of obtaining and determining the edge addition selection probability for each community pair based on the separation degree, attribute similarity, and the number of nodes in each community within the community pair includes: The attribute similarity of each community pair is obtained by the following formula: the cosine similarity between the average attribute vectors of the two communities within a community pair is taken as the attribute similarity of the community pair.

3. The method for hiding network communities based on user attribute features according to claim 1, characterized in that, The process of obtaining and determining the edge addition selection probability for each community pair based on the separation degree, attribute similarity, and the number of nodes in each community within the community pair includes: The edge addition selection probability for each community pair is obtained using the following formula: P J（ij） =|X i ||X j |×S(X i ,X j )× Among them, P J（ij） For the community to (X) i X j Add selection probability to the edge of X; i X is the community numbered i in the online community; j For community number j in the online community; |X i |For X i The number of nodes; |X j |For X j The number of nodes; S(X) i X j ) for the community to (X) i X j The resolution of ) For the community to (X) i X j The similarity of attributes.

4. The method for hiding network communities based on user attribute features according to claim 1, characterized in that, The process of obtaining and determining the edge deletion selection probability for each community based on its cohesion, attribute uniqueness, and number of node pairs includes: The attribute uniqueness of each community is obtained as follows: the cosine similarity between the average attribute vector of all communities in the network community and the average attribute vector of the current community is used as the attribute uniqueness of the current community.

5. The method for hiding network communities based on user attribute features according to claim 1, characterized in that, The process of obtaining and determining the edge deletion selection probability for each community based on its cohesion, attribute uniqueness, and number of node pairs includes: The edge deletion selection probability of each community is obtained by the following formula: P S（i） =C(X i ) ×|X i ||X i -1| Among them, P S（i） For X i The edge deletion selection probability; X i C(X) is the community numbered i in the online community. i ) is X i cohesion; For the uniqueness of community attributes; |X i ||X i -1| is twice the number of node pairs in the community.

6. The method for hiding network communities based on user attribute features according to claim 1, characterized in that, When adding edges to the non-edges with the smallest product of the sum of node degrees and (1 - attribute similarity) in the target community, the following conditions must be met: Select the non-edges that minimize the increase in the external degree score of the network community from the product of the sum of the node degrees of the target community and (1 - attribute similarity).

7. The method for hiding network communities based on user attribute features according to claim 1, characterized in that, When selecting and deleting edges in the target community that have the largest product of the sum of node degrees and (1 - attribute similarity), the following conditions apply: Among the edges in the target community, select the one that maximizes the increase in the external degree score of the network community by multiplying the sum of node degrees by (1 - attribute similarity). Then delete the edge.

8. A network community hiding system based on user attribute characteristics, characterized in that, include: The data acquisition module is used to acquire the graph structure, community information, and node attribute information of the network community. The budget allocation module is used to obtain the edge addition selection probability of each community pair based on the graph structure, community information and node attribute information of the network community. It also obtains the edge addition selection probability of each community pair based on the separation degree, attribute similarity and the number of nodes in each community within the community pair. For each community within the network community, obtain and based on the cohesion, attribute uniqueness, and number of node pairs of each community, obtain the edge deletion selection probability of each community; The edge addition module is used to randomly select community pairs as target community pairs based on the edge addition selection probability of each community pair, and select the non-edges with the smallest product of the sum of node degrees and (1-attribute similarity) in the target community pair for edge addition; The edge deletion module is used to randomly select a community as the target community based on the edge deletion selection probability of each community, and select the edge with the largest sum of node degree and (1-attribute similarity) product in the target community for edge deletion.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the network community hiding method based on user attribute features as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the network community hiding method based on user attribute features as described in any one of claims 1 to 7.

Images