Node influence maximization method based on adaptive propagation in social hypernetwork

Abstract

The present invention relates to a method for maximizing node influence based on adaptive propagation in a social super network. The method comprises: constructing a propagation weighted hypergraph according to the social super network topology, rumor propagation pathways and user information, and obtaining a user seed set that propagates rumor information according to the propagation weighted hypergraph. Constructing an adaptive propagation model. Obtaining the propagation influence of each user seed node according to a preset adaptive propagation evaluation system and propagation duration, screening the user seed set according to the propagation influence, and obtaining a candidate user set. Estimate the expected value of candidate user nodes in the candidate user set actively propagating rumors to user seed nodes in one stage, maximizing the rumor propagation strength through a heuristic search method, and obtaining the user seed node corresponding to the maximum rumor propagation strength as a key user. The use of this method can improve the search accuracy of key users that propagate rumor information while significantly reducing time overhead.

Description

Technical Field

[0001] The present invention relates to the technical field of maximizing the influence of hypernetwork nodes, and in particular to a method for maximizing the influence of nodes based on adaptive propagation in a social hypernetwork. Background Art

[0002] With the development of Internet information dissemination technology, more and more social platforms, e-commerce platforms and media information platforms have emerged. Users spread rumor information and advertising information on the Internet through targeted delivery, group chat forwarding and other methods. At present, there are many technologies and tools for analyzing social networks, such as graph theory algorithms, social network analysis software, etc., which use models to predict the spread of rumors to extract key information from aspects such as user personalization characteristics, information content characteristics and social network structure, which can help identify key nodes and groups in social networks. However, it is possible that a user spreads rumor information in a group chat, and there are several users in this group chat who are direct friends of the user. In terms of identifying the key users of rumor information dissemination, there are information conflicts and overlaps of key users in the dissemination process. It is impossible to extract information for the dissemination process, and it is impossible to reflect the new dissemination patterns and user behaviors in the network, resulting in low efficiency and accuracy in finding the initiators of spreading rumor information. Summary of the invention

[0003] Based on this, it is necessary to provide a node influence maximization method based on adaptive propagation in a social supernetwork, which can improve the search efficiency and accuracy of key users who spread rumor information, in response to the above technical problems.

[0004] A method for maximizing node influence based on adaptive propagation in a social super network, the method comprising:

[0005] A propagation weighted hypergraph is constructed according to the social hypernetwork topology, rumor propagation pathways and user information, and a user seed set that propagates rumor information is obtained according to the propagation weighted hypergraph.

[0006] An adaptive propagation model is constructed based on the user seed set and the propagation type of each user seed node.

[0007] The adaptive propagation model obtains the propagation influence of each user seed node according to the preset adaptive propagation evaluation system and propagation duration, and screens the user seed set according to the propagation influence to obtain the candidate user set.

[0008] Estimate the expected value of each candidate user node in the candidate user set actively spreading rumors to the user seed node in one stage to obtain the rumor propagation strength. After maximizing the rumor propagation strength through a heuristic search method, obtain the user seed node corresponding to the maximum value of the rumor propagation strength as the key user.

[0009] A device for maximizing node influence based on adaptive propagation in a social super network, the device comprising:

[0010] A user seed set acquisition module is used to construct a propagation weighted hypergraph based on the social hypernetwork topology, rumor propagation pathways, and user information, and to acquire a user seed set that propagates rumor information based on the propagation weighted hypergraph;

[0011] The adaptive propagation model building module is used to build an adaptive propagation model according to the user seed set and the propagation type of each user seed node.

[0012] The candidate user set acquisition module is used for the adaptive propagation model to obtain the propagation influence of each user seed node according to the preset adaptive propagation evaluation system and propagation duration, and to filter the user seed set according to the propagation influence to obtain the candidate user set;

[0013] The communication influence maximization module is used to estimate the expected value of each candidate user node in the candidate user set actively spreading rumors to the user seed node in one stage, and obtain the rumor propagation strength. After maximizing the rumor propagation strength through the heuristic search method, the user seed node corresponding to the maximum value of the rumor propagation strength is obtained as the key user.

[0014] A computer device comprises a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the following steps are implemented:

[0015] A propagation weighted hypergraph is constructed according to the social hypernetwork topology, rumor propagation pathways and user information, and a user seed set that propagates rumor information is obtained according to the propagation weighted hypergraph.

[0016] An adaptive propagation model is constructed based on the user seed set and the propagation type of each user seed node.

[0017] The adaptive propagation model obtains the propagation influence of each user seed node according to the preset adaptive propagation evaluation system and propagation duration, and screens the user seed set according to the propagation influence to obtain the candidate user set.

[0018] Estimate the expected value of each candidate user node in the candidate user set actively spreading rumors to the user seed node in one stage to obtain the rumor propagation strength. After maximizing the rumor propagation strength through a heuristic search method, obtain the user seed node corresponding to the maximum value of the rumor propagation strength as the key user.

[0019] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the following steps:

[0020] A propagation weighted hypergraph is constructed according to the social hypernetwork topology, rumor propagation pathways and user information, and a user seed set that propagates rumor information is obtained according to the propagation weighted hypergraph.

[0021] An adaptive propagation model is constructed based on the user seed set and the propagation type of each user seed node.

[0022] The adaptive propagation model obtains the propagation influence of each user seed node according to the preset adaptive propagation evaluation system and propagation duration, and screens the user seed set according to the propagation influence to obtain the candidate user set.

[0023] Estimate the expected value of each candidate user node in the candidate user set actively spreading rumors to the user seed node in one stage to obtain the rumor propagation strength. After maximizing the rumor propagation strength through a heuristic search method, obtain the user seed node corresponding to the maximum value of the rumor propagation strength as the key user.

[0024] The above-mentioned method for maximizing node influence based on adaptive propagation in the social super network first obtains a set of candidate users who receive rumor information and identifies users who potentially receive rumor information. The user group that may receive rumor information can be determined by social network analysis methods, such as based on user behavior, interests, social relationships and other factors. Secondly, an adaptive propagation model is constructed, and the propagation impact evaluation function is combined with the propagation probability to estimate the probability of each user actively propagating rumor information. Furthermore, the propagation impact evaluation function is used to obtain the expected value of each candidate user's active propagation, considering factors such as the user's influence, position in the social network, and previous propagation behavior, and as an evaluation parameter for estimating the adaptive propagation model, it can be updated and adjusted in real time according to the needs of the propagation model to improve the accuracy and effectiveness of the adaptive propagation model. Finally, by combining the expected propagation value and the heuristic search method, the user who is most likely to become the key propagator is identified, so as to achieve the improvement of the search accuracy of the key user who spreads the rumor information while reducing the time cost. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 A schematic diagram of a flow chart of a method for maximizing node influence based on adaptive propagation in a social super network in one embodiment;

[0026] Figure 2 It is a structural block diagram of a device for maximizing node influence based on adaptive propagation in a social super network in one embodiment;

[0027] Figure 3 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0028] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0029] In one embodiment, Figure 1 As shown, a method for maximizing node influence based on adaptive propagation in a social super network is provided, comprising the following steps:

[0030] Step 102, construct a propagation weighted hypergraph according to the social hypernetwork topology, rumor propagation pathways, and user information, and obtain a user seed set that propagates rumor information according to the propagation weighted hypergraph.

[0031] Step 104: construct an adaptive propagation model according to the user seed set and the propagation type of each user seed node.

[0032] Specifically, initially, the candidate user nodes in the network are set to the infected state infected(I), and the remaining non-candidate user nodes are in the susceptible state susceptible(S). In each time step t, for each candidate user node V in state i j , find the node V j All hyperedges E belonging to i = {ei1, ei2, ..., eim}, and the corresponding hyperedge weight is Wi = {wi1, wi2, ..., wim}. The hyperedge weight is normalized and converted into the probability P of the hyperedge being selected for propagation. ij ,Right now Then for each hyperedge e ij Generate a random number r between 0 and 1 ij , by comparing r ij With P ij To randomly determine whether the hyperedge is propagated. ij <=P ij When the super edge e ij Each s-state node contained in the selected propagation will be infected by the candidate user node with a probability of β Vi Infect.

[0033] Furthermore, the above process is terminated until a specific time step T is reached, where T is a control parameter, and an adaptive propagation model is constructed based on this propagation mode.

[0034] Step 106, the adaptive propagation model obtains the propagation influence of each user seed node according to the preset adaptive propagation evaluation system and propagation duration, and screens the user seed set according to the propagation influence to obtain the candidate user set.

[0035] Specifically, we construct a propagation impact evaluation function, namely (Expected Influence of One-hop Area, referred to as EIOA):

[0036]

[0037] Among them, EIOA(H) is the propagation expectation value of the candidate user set, H is the candidate user set, K is the number of candidate user nodes, M h is the set of nodes affected by the first-order propagation of the candidate user set H, R i For the network information interaction platform h Non-candidate user nodes in For the receiving node R i The set of neighbor nodes affected by the first-order propagation, R j for Candidate user nodes in P ji is the probability that the candidate user node spreads the rumor information to other non-candidate user nodes, The probability of selecting the nth hyperedge for a candidate user node to spread rumor information to non-candidate user nodes. Assume that the average degree of a candidate user node in the network is The average number of hyperedges is The time complexity of the multiplication term is The time complexity of accumulating items is The time complexity of the EIOA evaluation letter is The expected value of the expected influence of the candidate user set H in the propagation area in one stage can be calculated efficiently, thereby replacing multiple Monte Carlo simulations to obtain σ(H).

[0038] Step 108, estimate the expected value of each candidate user node in the candidate user set actively spreading rumors to the user seed node in one stage, obtain the rumor propagation strength, and after maximizing the rumor propagation strength through a heuristic search method, obtain the user seed node corresponding to the maximum value of the rumor propagation strength as the key user.

[0039] The preset heuristic search method can be EIOA-heuristic search or EIOA-avoid search.

[0040] Specifically, the EIOA-heuristic search method is selected to traverse each candidate user node in the weighted hypergraph and perform the EIOA function calculation on it. Then, the expected influence values ​​of the candidate user nodes are sorted in descending order and the corresponding node subscript sequence is returned. Finally, a specified number of candidate user nodes with high rankings are selected as key users.

[0041] Alternatively, you can use the EIOA-avoid search method:

[0042]

[0043] in, For the receiving node R i The set of neighbor nodes affected by first-order propagation, R q For the receiving node R i Non-candidate user nodes in the set of neighbor nodes affected by first-order propagation, For R q The corresponding candidate user nodes in the neighbor nodes affected by the first-order propagation have no effect on R q The probability of spreading rumor information, EIOA-avoid (R i ,H) is the candidate user node R i The expected value of the impact of propagation to non-candidate user nodes. ji is the probability that the candidate user node spreads the rumor information to other non-candidate user nodes, The probability of selecting the nth hyperedge for a candidate user node to spread rumor information to non-candidate user nodes.

[0044] Specifically, is the non-candidate user node R to be evaluated i The set of all neighbor nodes of one-stage propagation, R q is a non-candidate user node among these neighbor nodes. When , the expected value of its propagation probability is the same as that of the EIOA function. Based on the EIOA function, you also need to multiply by R q The non-candidate node pairs V in the neighbor nodes of the one-stage propagation m Based on this, the overlap of influence brought by key users determined by EIOA-heuristic can be avoided to a certain extent, and the search quality and accuracy of key user sets can be improved.

[0045] In the above-mentioned method for maximizing node influence based on adaptive propagation in social supernetwork, first, a set of candidate users who receive rumor information is obtained to identify potential users who receive rumor information. The user group that may receive rumor information can be determined by social network analysis methods, such as based on user behavior, interests, social relationships and other factors. Secondly, an adaptive propagation model is constructed to combine the propagation impact evaluation function with the propagation probability to estimate the probability of each user actively propagating rumor information. Furthermore, the propagation impact evaluation function is used to obtain the expected value of each candidate user's active propagation, considering factors such as the user's influence, position in the social network, and previous propagation behavior, and as an evaluation parameter for estimating the adaptive propagation model, it can be updated and adjusted in real time according to the needs of the propagation model to improve the accuracy and effectiveness of the adaptive propagation model. Finally, by combining the expected propagation value and the heuristic search method, the user who is most likely to become the key propagator is identified, so as to achieve the improvement of the search accuracy of the key user who spreads the rumor information while reducing the time cost.

[0046] In one embodiment, a propagation hypergraph is constructed based on the social super network topology as the hypergraph network topology, the rumor propagation path as the hyperedge, and the user information as the user seed node. After weighting each hyperedge in the propagation hypergraph, a propagation weighted hypergraph is obtained, and a user seed set that propagates rumor information in the social super network is obtained based on the propagation weighted hypergraph. The rumor propagation path includes: propagation type, number of information interactions of user seed nodes, and propagation frequency between user seed nodes, wherein the propagation type includes: global propagation, local propagation, and other propagations. User information includes: user node propagation status, user node information, and a set of user node information of all neighbors of the user node.

[0047] In one of the embodiments, an adaptive propagation model of a propagation weighted hypergraph is constructed based on the rumor propagation probability that the propagation state of a user seed node in a user seed set changes from a susceptible state to an infected state within a propagation time.

[0048] In one embodiment, the preset adaptive propagation evaluation system:

[0049]

[0050] Among them, EIOA(H) is the propagation expectation value of the candidate user set, H is the candidate user set, K is the number of candidate user nodes, M h is the set of nodes affected by the first-order propagation of the candidate user set H, R i For the network information interaction platform h Non-candidate user nodes in For the receiving node R i The set of neighbor nodes affected by the first-order propagation, R j for Candidate user nodes in P ji is the probability that the candidate user node spreads the rumor information to other non-candidate user nodes, The probability of selecting the nth hyperedge for a candidate user node to spread rumor information to non-candidate user nodes.

[0051] It is worth mentioning that the adaptive propagation model comprehensively considers that for the daily key users' behavior of spreading rumor information, there may be directed and fixed-point propagation between users from the same network platform, or there may be information propagation between users and group chats, enterprises and other multi-person circles. This propagation behavior is modeled separately in the local structure and dynamic propagation process. The point-to-point propagation process generates a local propagation feature model, and the point-to-surface (i.e., a set formed by multiple points) propagation process generates a global propagation feature model, which is represented by a hypergraph data structure, and also includes other propagation features that are not global propagation features or local propagation features.

[0052] In one embodiment, the adaptive propagation model obtains the hyperedge set of all neighboring user nodes corresponding to each user seed node and the user seed node according to the preset adaptive propagation evaluation system and propagation duration. Each hyperedge in the hyperedge weight set is weighted and normalized to obtain the propagation probability of each hyperedge. A random probability parameter is generated for each hyperedge in the hyperedge weight set. The propagation probability is compared with the random probability parameter. If the propagation probability is not less than the random probability parameter, all user seed nodes in the susceptible state contained in the hyperedge corresponding to the propagation probability will be selected as candidate user nodes for propagating rumor information after the next propagation duration ends, and a candidate user set is obtained. Otherwise, all user seed nodes in the susceptible state contained in the hyperedge corresponding to the propagation probability will be selected as candidate user nodes for propagating rumor information according to the random probability parameter after the next propagation duration ends, and a candidate user set is obtained.

[0053] In one embodiment,

[0054]

[0055] in, For the receiving node Ri The set of neighbor nodes affected by first-order propagation, Rq For the receiving node Ri Non-candidate user nodes in the set of neighbor nodes affected by first-order propagation, For R q The corresponding candidate user nodes in the neighbor nodes affected by the first-order propagation have no effect on R q The probability of spreading rumor information, EIOA-avoid (R i ,H) is the candidate user node Ri The expected value of the impact of propagation to non-candidate user nodes. ji is the probability that the candidate user node spreads the rumor information to other non-candidate user nodes, The probability of selecting the nth hyperedge for a candidate user node to spread rumor information to non-candidate user nodes.

[0056] It is worth noting that the EIOA-avoid search method is essentially a new heuristic evaluation indicator that considers the two-stage propagation range of the node, which can avoid the influence overlap caused by determining the key user set through the EIOA-heuristic to a certain extent, and further improve the search quality of the seed set. Among them, if a node is already a seed node, it needs to be punished. In this embodiment, the penalty value is set to -100 as the influence value of the node, which can also be any negative number.

[0057] In addition, the greedy strategy EIOA-greedy can also be used. Each time when searching for the next candidate user node, all non-candidate user nodes are greedily traversed, and they are added to the candidate user set respectively, and each new candidate user set is evaluated by the EIOA function. The candidate node corresponding to the maximum evaluation value is selected as the new key user node. The next key user node is always searched according to this process until the set number of key user sets is met. The time complexity of the algorithm is Based on EIOA, three search methods are designed to find the key users with the highest propagation probability. From a theoretical analysis, EIOA-heuristic is the simplest and most direct, and can quickly find multiple key users, but the accuracy may be slightly lower. EIOA-avoid can eliminate the influence overlap caused by EIOA-heuristic to a certain extent, and will not increase too much time overhead. EIOA-greedy can completely avoid the deviation caused by the overlap of node influence within the one-stage propagation range, but the time complexity will be greatly increased.

[0058] It should be understood that although Figure 1 The steps in the flowchart are shown in sequence as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified in this document, there is no strict order restriction for the execution of these steps, and these steps can be executed in other orders. Moreover, Figure 1 At least part of the steps may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least part of the sub-steps or stages of other steps.

[0059] In one embodiment, Figure 2 As shown, a node influence maximization device based on adaptive propagation in a social super network is provided, comprising: a user seed set acquisition module 202, an adaptive propagation model construction module 204, a candidate user set acquisition module 206 and a propagation influence maximization module 208, wherein:

[0060] The user seed set acquisition module 202 is used to construct a propagation weighted hypergraph according to the social hypernetwork topology, rumor propagation pathways and user information, and to acquire a user seed set for propagating rumor information according to the propagation weighted hypergraph.

[0061] The adaptive propagation model building module 204 is used to build an adaptive propagation model according to the user seed set and the propagation type of each user seed node.

[0062] The candidate user set acquisition module 206 is used for the adaptive propagation model to acquire the propagation influence of each user seed node according to the preset adaptive propagation evaluation system and propagation duration, and to filter the user seed set according to the propagation influence to obtain the candidate user set.

[0063] The communication influence maximization module 208 is used to estimate the expected value of each candidate user node in the candidate user set actively spreading rumors to the user seed node in a stage, obtain the rumor propagation strength, and after maximizing the rumor propagation strength through a heuristic search method, obtain the user seed node corresponding to the maximum value of the rumor propagation strength as the key user.

[0064] In one embodiment, the communication influence maximization module is also used to

[0065]

[0066] in, For the receiving node R i The set of neighbor nodes affected by first-order propagation, R q For the receiving node R i Non-candidate user nodes in the set of neighbor nodes affected by first-order propagation,

[0067] For R q The corresponding candidate user nodes in the neighbor nodes affected by the first-order propagation have no effect on R q The probability of spreading rumor information, EIOA-avoid (R i ,H) is the candidate user node R i The expected value of the impact of propagation to non-candidate user nodes. ji is the probability that the candidate user node spreads the rumor information to other non-candidate user nodes, The probability of selecting the nth hyperedge for a candidate user node to spread rumor information to non-candidate user nodes.

[0068] For the specific definition of the device for maximizing node influence based on adaptive propagation in a social super network, please refer to the definition of the method for maximizing node influence based on adaptive propagation in a social super network mentioned above, which will not be repeated here. Each module in the above-mentioned device for maximizing node influence based on adaptive propagation in a social super network can be implemented in whole or in part by software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0069] In one embodiment, a computer device is provided. The computer device may be a terminal, and its internal structure diagram may be as follows: Figure 3 As shown. The computer device includes a processor, a memory, a network interface, a display screen and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a method for maximizing the influence of nodes based on adaptive propagation in a social super network is implemented. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covered on the display screen, or a button, trackball or touchpad set on the computer device housing, or an external keyboard, touchpad or mouse, etc.

[0070] Those skilled in the art will understand that Figure 2-3 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present invention, and does not constitute a limitation on the computer device to which the solution of the present invention is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0071] In one embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the following steps are implemented:

[0072] A propagation weighted hypergraph is constructed according to the social hypernetwork topology, rumor propagation pathways and user information, and a user seed set that propagates rumor information is obtained according to the propagation weighted hypergraph.

[0073] An adaptive propagation model is constructed based on the user seed set and the propagation type of each user seed node.

[0074] The adaptive propagation model obtains the propagation influence of each user seed node according to the preset adaptive propagation evaluation system and propagation duration, and screens the user seed set according to the propagation influence to obtain the candidate user set.

[0075] Estimate the expected value of each candidate user node in the candidate user set actively spreading rumors to the user seed node in one stage to obtain the rumor propagation strength. After maximizing the rumor propagation strength through a heuristic search method, obtain the user seed node corresponding to the maximum value of the rumor propagation strength as the key user.

[0076] In one embodiment, when the processor executes the computer program, the following steps are further implemented: a preset adaptive propagation evaluation system:

[0077]

[0078] Among them, EIOA(H) is the propagation expectation value of the candidate user set, H is the candidate user set, K is the number of candidate user nodes, M h is the set of nodes affected by the first-order propagation of the candidate user set H, R i For the network information interaction platform h Non-candidate user nodes in For the receiving node R i The set of neighbor nodes affected by the first-order propagation, R j for Candidate user nodes in P ji is the probability that the candidate user node spreads the rumor information to other non-candidate user nodes, The probability of selecting the nth hyperedge for a candidate user node to spread rumor information to non-candidate user nodes.

[0079] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0080]

[0081] in, For the receiving node R i The set of neighbor nodes affected by first-order propagation, R q For the receiving node R i Non-candidate user nodes in the set of neighbor nodes affected by first-order propagation, For R q The corresponding candidate user nodes in the neighbor nodes affected by the first-order propagation have no effect on R qThe probability of spreading rumor information, EIOA-avoid (R i ,H) is the candidate user node R i The expected value of the impact of propagation to non-candidate user nodes. ji is the probability that the candidate user node spreads the rumor information to other non-candidate user nodes, The probability of selecting the nth hyperedge for a candidate user node to spread rumor information to non-candidate user nodes.

[0082] In one embodiment, a computer readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:

[0083] A propagation weighted hypergraph is constructed according to the social hypernetwork topology, rumor propagation pathways and user information, and a user seed set that propagates rumor information is obtained according to the propagation weighted hypergraph.

[0084] An adaptive propagation model is constructed based on the user seed set and the propagation type of each user seed node.

[0085] The adaptive propagation model obtains the propagation influence of each user seed node according to the preset adaptive propagation evaluation system and propagation duration, and screens the user seed set according to the propagation influence to obtain the candidate user set.

[0086] Estimate the expected value of each candidate user node in the candidate user set actively spreading rumors to the user seed node in one stage to obtain the rumor propagation strength. After maximizing the rumor propagation strength through a heuristic search method, obtain the user seed node corresponding to the maximum value of the rumor propagation strength as the key user.

[0087] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented: a preset adaptive propagation evaluation system:

[0088]

[0089] Among them, EIOA(H) is the propagation expectation value of the candidate user set, H is the candidate user set, K is the number of candidate user nodes, M h is the set of nodes affected by the first-order propagation of the candidate user set H, R i For the network information interaction platform h Non-candidate user nodes in For the receiving node R i The set of neighbor nodes affected by the first-order propagation, R j for Candidate user nodes in P ji is the probability that the candidate user node spreads the rumor information to other non-candidate user nodes, The probability of selecting the nth hyperedge for a candidate user node to spread rumor information to non-candidate user nodes.

[0090] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0091]

[0092] in, For the receiving node R i The set of neighbor nodes affected by first-order propagation, R q For the receiving node R i Non-candidate user nodes in the set of neighbor nodes affected by first-order propagation, For R q The corresponding candidate user nodes in the neighbor nodes affected by the first-order propagation have no effect on R q The probability of spreading rumor information, EIOA-avoid (R i ,H) is the candidate user node R i The expected value of the impact of propagation to non-candidate user nodes. ji is the probability that the candidate user node spreads the rumor information to other non-candidate user nodes, The probability of selecting the nth hyperedge for a candidate user node to spread rumor information to non-candidate user nodes.

[0093] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided by the present invention can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0094] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0095] The above-mentioned embodiments only express several implementation methods of the present invention, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.

Claims

1. A method for maximizing node influence based on adaptive propagation in a social hypernetwork, characterized in that: The method comprises: Constructing a propagation weighted hypergraph according to the social hypernetwork topology, rumor propagation pathways, and user information, and obtaining a user seed set that propagates rumor information according to the propagation weighted hypergraph; An adaptive propagation model is constructed according to the propagation type of the user seed set and each user seed node. Specifically, the candidate user nodes in the network are set to the infected state, and the remaining non-candidate user nodes are in the susceptible state. In each time step t, all hyperedges E to which each candidate user node in the infected state belongs are found. i ={ei1, ei2, ..., eim}, the corresponding hyperedge weight is Wi ={wi1, wi2, ..., wim}; the hyperedge weight is normalized and converted into the probability of the corresponding hyperedge being selected for propagation; for each hyperedge e ij Generate a random number r between 0 and 1 ij , by comparing the random number r ij The probability of the hyperedge being selected for propagation randomly determines whether the hyperedge is propagated; when the random number r i j is not greater than the probability P of the hyperedge being selected for propagation ij When the hyperedge e ij The selected nodes are propagated, and each susceptible node contained therein is infected by the candidate user node with a preset infection probability; until the preset time step is reached, the adaptive propagation is terminated; Adaptive Communication Evaluation System: Among them, EIOA(H) is the propagation expectation value of the candidate user set, H is the candidate user set, K is the number of candidate user nodes, M h is the set of nodes affected by the first-order propagation of the candidate user set H, R i For the network information interaction platform h Non-candidate user nodes in For the receiving node R i The set of neighbor nodes affected by the first-order propagation, R j for Candidate user nodes in P ji is the probability that the candidate user node spreads the rumor information to other non-candidate user nodes, The probability of selecting the nth hyperedge for a candidate user node to propagate rumor information to non-candidate user nodes; The adaptive propagation model obtains the propagation influence of each user seed node according to a preset adaptive propagation evaluation system and propagation duration, and screens the user seed set according to the propagation influence to obtain a candidate user set; Estimate the expected value of each candidate user node in the candidate user set actively spreading rumors to the user seed node in a stage to obtain the rumor propagation strength. After maximizing the rumor propagation strength through a heuristic search method, obtain the user seed node corresponding to the maximum rumor propagation strength as the key user.

2. The method according to claim 1, characterized in that A propagation weighted hypergraph is constructed according to the social hypernetwork topology, rumor propagation pathways, and user information, and a user seed set of propagation information is obtained according to the propagation weighted hypergraph, including: Construct a propagation hypergraph according to the social hypernetwork topology as the hypergraph network topology, the rumor propagation path as the hyperedge, and the user information as the user seed node, weight each hyperedge in the propagation hypergraph to obtain a propagation weighted hypergraph, and obtain a user seed set that propagates rumor information in the social hypernetwork according to the propagation weighted hypergraph; The rumor propagation pathway includes: propagation type, number of information interactions between user seed nodes, and propagation frequency between user seed nodes, wherein the propagation type includes: global propagation, local propagation, and other propagations; The user information includes: user node propagation status, user node information, and a collection of information of all neighboring user nodes of the user node.

3. The method according to claim 2, characterized in that An adaptive propagation model is constructed according to the propagation type of the user seed set and each user seed node, including: An adaptive propagation model of the propagation weighted hypergraph is constructed according to the rumor propagation probability that the propagation state of the user seed nodes in the user seed set is converted from a susceptible state to an infected state within the propagation duration.

4. The method according to claim 3, characterized in that The adaptive propagation model obtains the propagation influence of each user seed node according to a preset adaptive propagation evaluation system and propagation duration, and screens the user seed set according to the propagation influence to obtain a candidate user set, including: The adaptive propagation model obtains a hyperedge set of each user seed node and all neighbor user nodes corresponding to the user seed node according to a preset adaptive propagation evaluation system and propagation duration; Each hyperedge in the hyperedge weight set is weighted and normalized to obtain the propagation probability of each hyperedge; Generating a random probability parameter for each hyperedge in the hyperedge weight set; Compare the propagation probability with the random probability parameter. If the propagation probability is not less than the random probability parameter, all user seed nodes in the susceptible state contained in the hyperedge corresponding to the propagation probability will be selected as candidate user nodes for propagating rumor information after the next propagation duration ends, and a candidate user set is obtained. Otherwise, all user seed nodes in the susceptible state contained in the hyperedge corresponding to the propagation probability will be selected as candidate user nodes for propagating rumor information according to the random probability parameter after the next propagation duration ends, and a candidate user set is obtained.

5. The method according to claim 4, characterized in that Estimate the expected value of each candidate user node in the candidate user set actively spreading rumors to the user seed node in a stage to obtain the rumor propagation strength. After maximizing the rumor propagation strength through a heuristic search method, obtain the user seed node corresponding to the maximum rumor propagation strength as the key user, including: in, For the receiving node R i The set of neighbor nodes affected by first-order propagation, R q For the receiving node R i Non-candidate user nodes in the set of neighbor nodes affected by first-order propagation, For R q The corresponding candidate user nodes in the neighbor nodes affected by the first-order propagation have no effect on R q The probability of spreading rumor information, EIOA-avoid (R i ,H) is the candidate user node R i The expected value of the impact of propagation to non-candidate user nodes.

6. A device for maximizing node influence based on adaptive propagation in a social super network, characterized in that: The device comprises: A user seed set acquisition module is used to construct a propagation weighted hypergraph according to the social hypernetwork topology, rumor propagation pathways, and user information, and to acquire a user seed set for propagating rumor information according to the propagation weighted hypergraph; The adaptive propagation model construction module is used to construct an adaptive propagation model according to the propagation type of the user seed set and each user seed node. Specifically, the candidate user nodes in the network are set to the infected state, and the remaining non-candidate user nodes are in the susceptible state. In each time step t, all hyperedges E to which each candidate user node in the infected state belongs are found. i ={ei1, ei2, ..., eim}, the corresponding hyperedge weight is Wi ={wi1, wi2, ..., wim}; the hyperedge weight is normalized and converted into the probability of the corresponding hyperedge being selected for propagation; for each hyperedge e ij Generate a random number r between 0 and 1 ij , by comparing the random number r ij The probability of the hyperedge being selected for propagation randomly determines whether the hyperedge is propagated; when the random number r ij No greater than the probability P of the hyperedge being selected for propagation ij When the hyperedge e ij The selected nodes are infected by the candidate user nodes with a preset infection probability, and each susceptible node contained therein is infected by the candidate user nodes with a preset infection probability; until the preset time step is reached, the adaptive propagation is terminated; the adaptive propagation evaluation system: Among them, EIOA(H) is the propagation expectation value of the candidate user set, H is the candidate user set, K is the number of candidate user nodes, M h is the set of nodes affected by the first-order propagation of the candidate user set H, R i For the network information interaction platform h Non-candidate user nodes in For the receiving node R i The set of neighbor nodes affected by the first-order propagation, R j for Candidate user nodes in P ji is the probability that the candidate user node spreads the rumor information to other non-candidate user nodes, The probability of selecting the nth hyperedge for a candidate user node to propagate rumor information to non-candidate user nodes; A candidate user set acquisition module is used for the adaptive propagation model to acquire the propagation influence of each user seed node according to a preset adaptive propagation evaluation system and propagation duration, and to filter the user seed set according to the propagation influence to obtain a candidate user set; The communication influence maximization module is used to estimate the expected value of each candidate user node in the candidate user set actively spreading rumors to the user seed node in a stage, obtain the rumor propagation strength, and after maximizing the rumor propagation strength through a heuristic search method, obtain the user seed node corresponding to the maximum value of the rumor propagation strength as the key user.

7. The device according to claim 6, characterized in that The communication influence maximization module is also used in, For the receiving node Ri The set of neighbor nodes affected by first-order propagation, Rq For the receiving node Ri Non-candidate user nodes in the set of neighbor nodes affected by first-order propagation, For Rq The corresponding candidate user nodes in the neighbor nodes affected by the first-order propagation have no Rq The probability of spreading rumor information, EIOA-avoid (R i ,H) is the candidate user node Ri The expected value of the impact of propagation to non-candidate user nodes.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 5 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 5 are implemented.

Images