Network monitoring methods, devices, equipment, storage media, programs, and program products

By constructing a topic hypernetwork model and classifying topics using information released on network platforms, the problem of low reliability in existing network monitoring technologies is solved, and more accurate online public opinion monitoring is achieved.

CN118797185BActive Publication Date: 2026-11-17CHINA MOBILEHANGZHOUINFORMATION TECH CO LTD +1
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202410167800.6
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-02-05
Publication Date
2026-11-17
Estimated Expiration
2044-02-05

AI Technical Summary

Technical Problem

Existing methods for monitoring online public opinion can only monitor trending topics, leading to information loss and low reliability of online monitoring.

Method used

A topic hypernetwork model is constructed, which utilizes the release of relevant information on a network platform at multiple time periods. The topic hypernetwork model is used to classify topics, including a set of nodes, a set of node weights, and a set of hyperedges, to describe the relationships and dependencies between topics and to simulate the changing trends of topics over time.

Benefits of technology

This has enriched the types of topics monitored on the online platform, expanded the monitoring scope, and improved the accuracy of topic classification and the reliability of online monitoring.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118797185B_ABST
    Figure CN118797185B_ABST
Patent Text Reader

Abstract

This application discloses a network monitoring method, apparatus, device, storage medium, program, and program product, belonging to the field of network technology. The method includes: acquiring relevant information published by a network platform over T time periods; constructing T topic hypernetwork models corresponding to the T time periods using the published information, wherein the t-th topic hypernetwork model corresponding to the t-th time period includes a node set, a node weight set, and a hyperedge set; the node set includes f nodes corresponding to f topics; the node weight set includes f node weights corresponding to the f nodes; the hyperedge set includes r hyperedges corresponding to r topic sources, and each topic source's hyperedge includes n nodes corresponding to n topics published through that topic source; and classifying the topics in the network platform using the T topic hypernetwork models. This application can improve the reliability of network monitoring.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This application belongs to the field of network technology, and in particular relates to a network monitoring method, device, equipment, storage medium, program, and program product. Background Technology

[0002] With the widespread adoption of the internet, its role in guiding public opinion has become increasingly prominent. Online public opinion refers to various online opinions circulating about social events, characterized by low ignition points, suddenness, rapid dissemination, and strong interactivity, which to some extent impacts cybersecurity. Therefore, monitoring technologies for online public opinion are of great significance in online public opinion management.

[0003] Currently, the main methods for monitoring online public opinion include machine monitoring. Existing machine monitoring methods involve: first, segmenting the online public opinion information to be monitored into word groups to generate candidate topic groups; then, calculating the semantic similarity of words in the candidate topic groups and merging topics with high similarity as the topics to be monitored; subsequently, identifying keywords in the topics to be monitored that appear more frequently than a threshold as trending topics. While this method is simple and quick to calculate based on the statistical characteristics of words, it can only monitor trending topics, easily leading to information loss and resulting in low reliability of online monitoring. Summary of the Invention

[0004] This application provides a network monitoring method, apparatus, device, storage medium, program, and program product that can solve the problem of low reliability in network monitoring.

[0005] In a first aspect, embodiments of this application provide a network monitoring method, the method comprising:

[0006] Obtain relevant information published by a network platform over T time periods, wherein the relevant information published by the network platform in time period t includes: r topic sources included in the network platform in time period t, f topics published through the r topic sources in time period t, and the publication frequency of each of the f topics in time period t; T is a positive integer, t is a positive integer less than or equal to T, r is a positive integer, and f is an integer greater than 1;

[0007] Using the publication information from the T time periods, T topic hypernetwork models corresponding one-to-one with the T time periods are constructed. The t-th topic hypernetwork model corresponding to the t-th time period includes a node set, a node weight set, and a hyperedge set. The node set includes f nodes corresponding one-to-one with the f topics. The node weight set includes f node weights corresponding one-to-one with the f nodes, and the node weight of each node is determined based on the publication frequency of the topic corresponding to the node within the t-th time period. The hyperedge set includes r hyperedges corresponding one-to-one with the r topic sources, and the hyperedge corresponding to each topic source includes n nodes corresponding one-to-one with the n topics published through the topic source; n is a positive integer less than f.

[0008] The topics in the network platform are classified using the T topic hypernetwork models.

[0009] Secondly, embodiments of this application provide a network monitoring device, the device comprising:

[0010] The acquisition module is used to acquire relevant information published by the network platform in T time periods. The relevant information published by the network platform in the t-th time period includes: r topic sources included in the network platform in the t-th time period, f topics published through the r topic sources in the t-th time period, and the publication frequency of each of the f topics in the t-th time period; T is a positive integer, t is a positive integer less than or equal to T, r is a positive integer, and f is an integer greater than 1.

[0011] A construction module is used to construct T topic hypernetwork models corresponding one-to-one with the T time periods using the release-related information of the T time periods. The t-th topic hypernetwork model corresponding to the t-th time period includes a node set, a node weight set, and a hyperedge set. The node set includes f nodes corresponding one-to-one with the f topics. The node weight set includes f node weights corresponding one-to-one with the f nodes, and the node weight of each node is determined based on the release frequency of the topic corresponding to the node within the t-th time period. The hyperedge set includes r hyperedges corresponding one-to-one with the r topic sources, and the hyperedge corresponding to each topic source includes n nodes corresponding one-to-one with the n topics released through the topic source; n is a positive integer less than f.

[0012] The classification module is used to classify topics in the network platform using the T topic hypernetwork models.

[0013] Thirdly, embodiments of this application provide a network monitoring device, the device including: a processor and a memory storing computer program instructions; the processor executes the computer program instructions to implement the network monitoring method as described in the first aspect.

[0014] Fourthly, embodiments of this application provide a computer storage medium storing computer program instructions, which, when executed by a processor, implement the network monitoring method as described in the first aspect.

[0015] Fifthly, embodiments of this application provide a computer program including computer-readable code, which, when executed by a processor, implements the network monitoring method as described in the first aspect.

[0016] In a sixth aspect, embodiments of this application provide a computer program product in which instructions, when executed by a processor of an electronic device, cause the electronic device to perform the network monitoring method as described in the first aspect.

[0017] In this embodiment, T topic hypernetwork models corresponding one-to-one with the T time periods can be constructed using the information published by the network platform over T time periods. The topic hypernetwork model corresponding to the t-th time period includes a node set, a node weight set, and a hyperedge set. The node set includes f nodes corresponding to f topics published in the t-th time period; the node weight set includes f node weights corresponding to the f nodes; and the hyperedge set includes r hyperedges corresponding one-to-one with the r topic sources included in the network platform during the t-th time period. Each hyperedge corresponding to a topic source includes nodes corresponding to topics published through that topic source. In this way, the hierarchical relationships between topic sources and topics, as well as the co-occurrence relationships between topics, are transformed into a network structure. This allows the topic hypernetwork model corresponding to each time period to describe the relationships between topics within that time period, and the topic hypernetwork models corresponding to different time periods to describe the changing trends of topics over time. Thus, by using T topic hypernetwork models to classify topics in the network platform, the types of topics monitored in the network platform can be enriched, the monitoring scope of the network platform can be expanded, and the accuracy of topic classification can be improved, thereby enhancing the reliability of network monitoring. Attached Figure Description

[0018] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments of this application will be briefly introduced below. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is one of the flowcharts of the network monitoring method provided in the embodiments of this application;

[0020] Figure 2 This is a schematic diagram of the topic hypernetwork model provided in the embodiments of this application;

[0021] Figure 3 This is a schematic diagram of the fully connected graph of the topic hypernetwork model provided in the embodiments of this application;

[0022] Figure 4 This is the second flowchart of the network monitoring method provided in the embodiments of this application;

[0023] Figure 5 This is a structural diagram of the network monitoring device provided in the embodiments of this application;

[0024] Figure 6 This is a structural diagram of the network monitoring device provided in the embodiments of this application. Detailed Implementation

[0025] The features and exemplary embodiments of various aspects of this application will be described in detail below. To make the objectives, technical solutions, and advantages of this application clearer, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only intended to explain this application and not to limit it. For those skilled in the art, this application can be implemented without some of these specific details. The following description of the embodiments is merely to provide a better understanding of this application by illustrating examples.

[0026] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising..." does not exclude the presence of additional identical elements in the process, method, article, or apparatus that includes said element.

[0027] The network monitoring method provided in this application will be described in detail below with reference to the accompanying drawings and through some embodiments and application scenarios.

[0028] See Figure 1 , Figure 1 This is one of the flowcharts of the network monitoring method provided in the embodiments of this application. For example... Figure 1 As shown, a network monitoring method may include the following steps:

[0029] Step 101: Obtain the relevant information published by the network platform in T time periods, wherein the relevant information published by the network platform in the t-th time period includes: r topic sources included in the network platform in the t-th time period, f topics published through the r topic sources in the t-th time period, and the publication frequency of each of the f topics in the t-th time period; T is a positive integer, t is a positive integer less than or equal to T, r is a positive integer, and f is an integer greater than 1.

[0030] The r topic sources included in the network platform during time period t can include: topic sources that already existed on the network platform at the start time of time period t, and topic sources newly created on the network platform during time period t.

[0031] In this embodiment, the topic source can be a post, article, microblog, etc. The topic source itself includes content, and after a topic source is published, users can comment on it. The topic published through that topic source can be obtained by processing the content of the topic source and the comments within it.

[0032] This application does not limit the above-described processing methods for content / comments. In some embodiments, the content / comments can be segmented into words while removing punctuation and stop words to obtain the word segments. For each topic source, the word segments of the content and comments of the topic source can be counted to obtain at least one word segment and the statistical frequency of each word segment. Then, each word segment in the at least one word segment can be regarded as a topic, which are the topics published through the topic source. The statistical frequency of each word segment in the at least one word segment can be determined as the publication frequency of the topic corresponding to that word in the topic source.

[0033] The f topics published through r topic sources in the t-th time period can include: new topics published through r topic sources within the t-th time period.

[0034] For r topic sources that already exist on the network platform at the start time of time period t, the topic published through that topic source in time period t can be determined as follows: obtain the new comments published through that topic source within time period t; process these new comments to obtain the topic published through that topic source in time period t.

[0035] For a new topic source created on the network platform within the time period t out of r topic sources, since the topic source was newly created within the time period t, the topic published through the topic source in the time period t can be determined in the following way: obtain all comments published through the topic source; process the content of the topic source and all comments of the topic source to obtain the topic published through the topic source in the time period t.

[0036] The posting frequency of each topic within time t can be represented as the sum of the statistical counts of each topic source for the corresponding word segment within time t.

[0037] In some embodiments, the information to be published at time t may further include: the user attention of each of the r topic sources in the time period t.

[0038] The user engagement of a topic source during time period t can be determined based on at least one of the following: the number of new comments posted by the topic source during time period t; or the number of new views generated by the topic source during time period t. It is worth noting that if the topic source is newly created during time period t, its comment count includes the content of the topic source, i.e., its comment count is its actual number of comments + 1. In some implementations, the user engagement of a topic source during time period t can be the sum of the above two, but it is not limited to this.

[0039] The T time periods can be specified time periods for the online platform, and they can be continuous or non-contiguous. Understandably, the topics, published topics, and related quantities may differ across different time periods.

[0040] Step 102: Using the publication information of the T time periods, construct T topic hypernetwork models corresponding one-to-one with the T time periods. The topic hypernetwork model corresponding to the t-th time period includes a node set, a node weight set, and a hyperedge set. The node set includes f nodes corresponding one-to-one with the f topics. The node weight set includes f node weights corresponding one-to-one with the f nodes, and the node weight of each node is determined based on the publication frequency of the topic corresponding to the node within the t-th time period. The hyperedge set includes r hyperedges corresponding one-to-one with the r topic sources, and the hyperedge corresponding to each topic source includes n nodes corresponding one-to-one with the n topics published through the topic source; n is a positive integer less than f.

[0041] In some embodiments, the t-th topic hypernetwork model may further include a set of leading weights, wherein the set of leading weights includes r leading weights that correspond one-to-one with the r leading edges, and the leading weight of each leading edge is determined based on the user attention of the leading edge in the t-th time period.

[0042] In practice, the relevant information published within a certain time period can be used to construct a topic hypernetwork model corresponding to that time period. This transforms the published relevant information within that time period into a network structure, enabling the topic network model corresponding to that time period to describe the correlation between various topics within that time period, as well as the subordinate relationship between topics and topic sources.

[0043] In this embodiment of the application, in order to discover the changing trend of topics on the network platform over time, the relevant information published in T time periods can be used to construct topic hypernetwork models corresponding to the T time periods respectively, so as to obtain a hypernetwork evolutionary network (HDEN) model that can simulate the time-varying state of topics. HDEN includes T topic hypernetwork models, which can be represented by formula (1):

[0044] HDEN = HN t =(V,Q,E,W) t ,t=1,2,...T (1)

[0045] Among them, HN t Let (V, Q, E, W) represent the hypernetwork model for the topic at time t. t In this context, V represents the set of nodes in the t-th topic hypernetwork model, including f nodes, which can be represented as: V = {v1, v2, ..., v...} f}; Q represents the set of node weights in the t-th topic hypernetwork model, including the node weights corresponding to f nodes, which can be represented as: E represents the set of hyperedges in the t-th topic hypernetwork model, including r hyperedges, which can be represented as: E = {e1, e2, ..., e} r}, each super edge It consists of n topics; W represents the set of hyperedge weights of the t-th topic hypernetwork model, including the hyperedge weights corresponding to r hyperedges, which can be expressed as:

[0046] To facilitate understanding of the topic hypernetwork model, the following example illustrates the concept:

[0047] Suppose that the online platform creates 3 new posts in time period t, namely post 1, post 2 and post 3.

[0048] Post 1 contains 4 comments (including the original post) and has been viewed 10 times. Four topics were selected: Topic 1, Topic 2, Topic 3, and Topic 6. Post 2 contains 4 comments (including the original post) and has been viewed 18 times. Four topics were selected: Topic 4, Topic 5, Topic 3, and Topic 6. Post 3 contains 2 comments (including the original post) and has been viewed 2 times. Two topics were selected: Topic 7 and Topic 6.

[0049] Topic 1 has a total posting frequency of 1 in 3 posts, Topic 2 has a total posting frequency of 3 in 3 posts, Topic 3 has a total posting frequency of 2 in 3 posts, Topic 4 has a total posting frequency of 1 in 3 posts, Topic 5 has a total posting frequency of 2 in 3 posts, Topic 6 has a total posting frequency of 5 in 3 posts, and Topic 7 has a total posting frequency of 1 in 3 posts.

[0050] Therefore, the topic hypernetwork model built based on the above three posts can be as follows: Figure 2 As shown, in Figure 2 In the topic hypernetwork model, there are three hyperedges: e1 corresponding to post 1, e2 corresponding to post 2, and e3 corresponding to post 3.

[0051] like Figure 2 As shown, hyperedge e1 includes: node v1 corresponding to topic 1, node v2 corresponding to topic 2, node v3 corresponding to topic 3, and node v6 corresponding to topic 6. Hyperedge e2 includes: node v4 corresponding to topic 4, node v5 corresponding to topic 5, node v3 corresponding to topic 3, and node v6 corresponding to topic 6. Hyperedge e3 includes: node v7 corresponding to topic 7, and node v6 corresponding to topic 6.

[0052] The weight of the superedge e1 The weight of the superedge e2 Hyperedge weight of hyperedge e3 Node weight of node v1 The node weight of node v2 is 1. The node weight of node v3 is 3. The node weight of node v4 is 2. The node weight of node v5 is 1. The node weight q of node v6 is 2. v6 The node weight q of node v7 is 5. v1 The value is 1.

[0053] The common nodes included in different hyperedges can be called the common adjacency nodes of these hyperedges. Figure 2 In the given equation, nodes v3 and v6 co-occur in both hyperedge e1 and hyperedge e2; therefore, nodes v3 and v6 are common adjacencies of hyperedge e1 and hyperedge e2. Node v6 co-occurs in hyperedge e1, hyperedge e2, and hyperedge e3; therefore, node v6 is a common adjacency of hyperedge e1, hyperedge e2, and hyperedge e3.

[0054] Step 103: Use the T topic hypernetwork models to classify the topics in the network platform.

[0055] In this embodiment, after obtaining T topic hypernetwork models, the topics on the network platform can be classified using these models. Since each of the T topic hypernetwork models can describe the relationships between topics within a time period, and the hierarchical relationship between topics and their sources, and since topic hypernetwork models for different time periods can describe the time-varying state of topics, the accuracy of topic classification can be improved, thereby enhancing the reliability of network monitoring.

[0056] In some embodiments, the categorization results of a topic may include at least one of the following:

[0057] Hot topics: Discussion volume;

[0058] Non-hot topics: These topics have consistently low popularity and have no chance of becoming hot topics in the near future.

[0059] A topic in decline: once highly popular but gradually fading in popularity over time;

[0060] Popular related topics: These topics are strongly related to popular topics and have the potential to become popular topics as public attention shifts.

[0061] Furthermore, trending topics can be any of the following:

[0062] Mature trending topics: These topics have high discussion volume, numerous related topics, and a clear centrality;

[0063] Popular topics with high growth potential: These topics generate significant discussion and experience explosive and continuous growth over time.

[0064] In some embodiments, after determining the classification results of each topic, corresponding topic tracking or processing measures can be taken for each topic under each classification result. The specific measures can be set according to actual needs, and this application embodiment does not limit this.

[0065] The network monitoring method in this embodiment utilizes the information released by the network platform over T time periods to construct T topic hypernetwork models corresponding one-to-one with each of the T time periods. The topic hypernetwork model corresponding to the t-th time period includes a node set, a node weight set, and a hyperedge set. The node set includes f nodes corresponding to f topics released in the t-th time period; the node weight set includes f node weights corresponding to the f nodes; and the hyperedge set includes r hyperedges corresponding one-to-one with the r topic sources included in the network platform during the t-th time period. Each hyperedge corresponding to a topic source includes nodes corresponding to topics released through that topic source. In this way, the hierarchical relationships between topic sources and topics, as well as the co-occurrence relationships between topics, are transformed into a network structure. This allows the topic hypernetwork model corresponding to each time period to describe the relationships between topics within that time period, and the topic hypernetwork models corresponding to different time periods to describe the changing trends of topics over time. Thus, by using T topic hypernetwork models to classify topics in the network platform, the monitoring types of topics in the network platform can be enriched, the monitoring scope of the network platform can be expanded, and the accuracy of topic classification can be improved, thereby enhancing the reliability of network monitoring.

[0066] In practical applications, at least one of the T topic hypernetwork models can be used to classify topics in the network platform, as explained below.

[0067] In some embodiments, classifying topics in the network platform using the T topic hypernetwork models includes:

[0068] From the supernetwork model of topic T corresponding to time period T, obtain the weight of node T corresponding to the first node;

[0069] When the weight of the T-th node is greater than the weight threshold, the betweenness centrality and node superdegree corresponding to the first node are determined according to the T-th topic hypernetwork model. The betweenness centrality is used to characterize the contribution of the first topic corresponding to the first node to other topics; the node superdegree is used to characterize the number of hyperedges including the first node in the T-th topic hypernetwork model.

[0070] If the centrality of the intermediaries is greater than the centrality threshold and the exponent degree of the nodes is greater than the exponent degree threshold, the first topic is determined to be a popular topic.

[0071] The first node is the node corresponding to the first topic. The first topic can be any uncategorized topic published on the network platform during the T-time period.

[0072] In this embodiment, the node weight corresponding to the first node, i.e., the T-th node weight, can be obtained from the T-th topic hypernetwork model corresponding to the T-th time period. Then, the T-th node weight is compared with a weight threshold, where the weight threshold can be used to measure the popularity of the node.

[0073] If the weight of node T is greater than the weight threshold, it means that node T is highly popular in time period t, and topic T can be considered a candidate topic for popular topics. Otherwise, it means that node T is not highly popular in time period t, and topic T can be abandoned as a candidate topic for popular topics.

[0074] To improve the accuracy of identifying trending topics, this embodiment further introduces betweenness centrality and node superdegree to classify topics.

[0075] In a topic hypernetwork model, the betweenness centrality of a node can be used to characterize the contribution of the topic corresponding to that node to other topics, and can be determined based on the correlation between that node and other nodes in the topic hypernetwork model.

[0076] The node superdegree corresponding to a node in the topic hypernetwork model can be used to characterize the number of hyperedges in the topic hypernetwork model that include that node. In some embodiments, the node superdegree corresponding to a node in the topic hypernetwork model can be the number of hyperedges in the topic hypernetwork model that include that node, which can be expressed by formula (2):

[0077]

[0078] Among them, C N (i) represents the degree of node i; θ j Indicate whether superedge j contains node i; if it does, θ j =1, otherwise θ j =0.

[0079] by Figure 2 Taking the topic hypernetwork model shown as an example, in Figure 2 In the given equation, hyperedges e1, e2, and e3 all contain node v6; therefore, the superdegree of node v6 is 3. Hyperedges e1 and e2 both contain node v3; therefore, the superdegree of node v3 is 2. Node v7 is contained only by hyperedge e3; therefore, the superdegree of node v7 is 1.

[0080] In practice, after selecting the first node as a candidate topic for hot topics, the intermediate centrality and node superdegree of the first node can be determined based on the supernetwork model of the T-th topic.

[0081] Next, the centrality of the first node is compared with the centrality threshold, and the exometry of the first node is compared with the exometry threshold.

[0082] The centrality threshold measures the centrality and importance of a topic; the excess threshold measures the frequency of a topic within its source. If the centrality of the first node is greater than the centrality threshold, the first topic has strong centrality and importance; conversely, if it is less, the first topic has lower centrality and importance. If the excess of the first node is greater than the excess threshold, the first topic appears frequently; conversely, if it is less, the first topic appears infrequently.

[0083] In this embodiment, if the betweenness centrality of the first node is greater than the centrality threshold, and the node superdegree of the first node is greater than the superdegree threshold, the first topic can be identified as a hot topic. In this case, the first topic has high popularity, high frequency of occurrence, and strong topic centrality and importance, and can be further identified as a mature hot topic. Otherwise, the first topic can be abandoned as a candidate topic for hot topics.

[0084] In this embodiment, the popularity of the first topic in the T-time period can be determined by comparing the node weight (i.e., the T-node weight) of the first node in the T-topic hypernetwork model corresponding to the T-time period with a weight threshold. If the T-node weight of the first node is greater than the weight threshold, the betweenness centrality and node superdegree of the first node determined according to the T-topic hypernetwork model can be further used to determine whether the first topic can be identified as a popular topic. This improves the accuracy of identifying popular topics.

[0085] In some embodiments, after obtaining the weight of the Tth node corresponding to the first node from the Tth topic hypernetwork model corresponding to the Tth time period, the method may further include:

[0086] Obtain the weight of the first node corresponding to the first node from the super network model of the topic corresponding to the time period T-1;

[0087] If both the weight of node T and the weight of node T-1 are less than or equal to the weight threshold, the first topic is determined to be a non-popular topic; or,

[0088] If the centrality of a node is less than or equal to the centrality threshold, or the node exponentiation is less than the exponentiation threshold, and the difference between the weight of the T-th node and the weight of the (T-1)-th node is greater than the difference threshold, then the first topic is determined to be a popular topic; or...

[0089] If the weight of node T is less than or equal to the weight threshold, but the weight of node T-1 is greater than the weight threshold, the first topic is determined to be a declining topic.

[0090] In this embodiment, if the weight of the T-th node corresponding to the first node is less than or equal to the weight threshold, or the centrality of the first node is less than or equal to the centrality threshold, or the superdegree of the first node is less than the superdegree threshold, the first topic can be monitored by combining the node weight in the supernetwork model of the T-1 topic corresponding to the first node in the T-1 time period, i.e., the T-1 node weight. The T-1 time period is the time period preceding the T-th time period among the T time periods.

[0091] In practice, if the weight of the Tth node corresponding to the first node is less than or equal to the weight threshold, the weight of the (T-1)th node corresponding to the first node can be compared with the weight threshold.

[0092] If the weight of node T-1 is also less than or equal to the weight threshold, it indicates that the popularity of the first topic remains low, and it can be identified as a non-popular topic.

[0093] If the weight of node T-1 is greater than the weight threshold, it indicates that the popularity of the first topic is declining, and it can be identified as a declining topic.

[0094] If the centrality of the betweenness corresponding to the first node is less than or equal to the centrality threshold, or if the superdegree of the node corresponding to the first node is less than the superdegree threshold, the difference between the weight of the (T-1)th node and the weight of the Tth node can be calculated and compared with the difference threshold. If the difference is greater than the difference threshold, it indicates that the popularity of the first topic is continuously increasing, and it can be identified as a hot topic. In this case, the first topic has high popularity, and its popularity continues to increase over time, so it can be further identified as a growing hot topic. If the difference is less than the difference threshold, it can be identified as a non-hot topic.

[0095] In this embodiment, the first topic is further monitored by combining the node weight of the first node in the supernetwork model of the topic corresponding to the first node in the time period T-1, i.e., the weight of the T-1 node, and the topic type of the first topic is determined. In this way, the reliability of topic monitoring can be further improved.

[0096] This application does not limit the method for determining the betweenness centrality. In some embodiments, determining the betweenness centrality and node superdegree corresponding to the first node based on the T-th topic hypernetwork model may include:

[0097] By fully connecting the nodes in each hyperedge of the T-th topic hypernetwork model, the fully connected graph corresponding to the T-th topic hypernetwork model is obtained.

[0098] The contribution of the first node to each node group is obtained by querying the fully connected graph. Each node group includes any two nodes in the T-th topic hypernetwork model other than the first node, and at least one node in each node group is different. The contribution of the first node to each node group is determined based on the shortest connection path between the two nodes in the node group in the fully connected graph.

[0099] The intermediary centrality of the first node is determined based on its contribution to each node group.

[0100] In this embodiment, for the T-th topic hypernetwork model, the nodes in each hyperedge can be fully connected to obtain the fully connected graph of the T-th topic hypernetwork model.

[0101] In practice, when performing a full connection on a node within a hyperedge, you can first select a node and connect it to all other nodes with straight lines. Then, select another node and connect it to all other nodes with straight lines. If a straight line connecting a node to another node intersects with an existing line, cancel the connection. Continue this process until all nodes have been traversed. For easier understanding, please refer to [link to documentation / reference]. Figure 3 , Figure 3 for Figure 2 The topic hypernetwork model shown is a fully connected graph. Figure 3 The fully connected graph shown is obtained by selecting nodes in ascending order of their labels and connecting them with other nodes by straight lines.

[0102] After obtaining the fully connected graph corresponding to the supernetwork model of topic T, the contribution of the first node to each node group can be obtained by querying the fully connected graph, thereby determining the contribution of the first topic to other topics.

[0103] In practice, we can first obtain the node groups corresponding to the first node; then, obtain the shortest connection path between the two nodes in each node group in the fully connected graph. The length of the connection path can be determined by the number of nodes it passes through. The more nodes the connection path passes through, the longer the connection path is, and vice versa. Then, based on the shortest connection path corresponding to each node group, we can determine the contribution of the first node to each connection group.

[0104] Assume the first node is Figure 3 If node v1 is in the list, then the node groups corresponding to node v2 include: v2 and v3, v2 and v4, v2 and v5, v2 and v6, v2 and v7, v3 and v4, v3 and v4, v3 and v5, v3 and v6, v3 and v7, v4 and v5, v4 and v6, v4 and v7, v5 and v6, v5 and v7, v6 and v7.

[0105] Taking the node group v2 and v7 as an example, the connection paths between them can include: v2→v1→v6→v7, v2→v3→v6→v7, v2→v3→v5→v6→v7, v2→v3→v4→v5→v6→v7, and v2→v3→v4→v5→v6→v7. The shortest paths between them include: v2→v1→v6→v7 and v2→v3→v6→v7.

[0106] After determining the shortest connection path for a certain group of nodes, the contribution of the first node to the connection group can be determined in ways that are not limited to the following.

[0107] In some implementations, querying the fully connected graph to obtain the contribution of the first node to each node group includes:

[0108] For each node group, query the fully connected graph to obtain g shortest connection paths between two nodes in the node group in the fully connected graph, and h shortest connection paths including the first node, where g is a positive integer and h is a positive integer less than or equal to g.

[0109] The ratio of h to g is determined as the contribution of the first node to the node group.

[0110] In this embodiment, after determining the shortest connection path g corresponding to a certain node group, the number h of paths including the first node in each shortest connection path of g can be determined. Then, the ratio of h to g is determined as the contribution of the first node to the node group. In this way, the contribution of the first node to the node group can accurately reflect the importance of the first topic as a bridge between the topics corresponding to the two nodes in the node group, thereby improving the accuracy of topic classification.

[0111] In other implementations, after obtaining the number of shortest paths including the first node for each node group, the number of shortest paths including the first node for each node group can be normalized, and the normalized result of the number of shortest paths including the first node for each node group can be determined as the contribution of the first node to the node group.

[0112] After obtaining the contribution of the first node to each node group, the sum of the contributions of the first node to each node group can be determined as the intermediary centrality of the first node. In this case, the intermediary centrality of the first node can be expressed by formula (3):

[0113]

[0114] Among them, C B (i) represents the intermediary centrality of node i; g jk This represents the number of shortest connection paths from node j to node k; gjk (i) represents the number of shortest connection paths from node j to node k that include node i.

[0115] Of course, in other implementations, the product or average of the contributions of the first node to each node group can also be used to determine the betweenness centrality of the first node. It is understood that the centrality threshold can be set differently depending on the method of determining betweenness centrality, and the specific setting can be determined according to actual needs; this application does not limit this.

[0116] In this embodiment, by obtaining the fully connected graph corresponding to the supernetwork model of the Tth topic, the contribution of the first node to each node group is determined. Then, the intermediary centrality of the first node is determined by using the contribution of the first node to each node group. In this way, the intermediary centrality of the node can accurately reflect the contribution of the topic corresponding to the node to other topics, thereby improving the accuracy of topic classification.

[0117] In some embodiments, classifying topics in the network platform using the T topic hypernetwork models includes:

[0118] Determine at least two neighboring nodes of the second node in the supernetwork model of the Tth topic corresponding to the Tth time period, wherein the second topic corresponding to the second node is a hot topic;

[0119] For each of the at least two adjacent nodes, obtain the number of shared super edges between the adjacent node and the second node, wherein the shared super edges are the super edges in the T-th topic supernetwork model that include the adjacent node and the second node;

[0120] The third topic corresponding to the third node is determined as a popular related topic. The third node is the Nth neighboring node among the at least two neighboring nodes that has the most shared superedges with the second node, where N is a positive integer.

[0121] The second node is the node corresponding to the second topic, which is a trending topic.

[0122] In practical implementation, we can first query the adjacent nodes of the second node in the T-th topic hypernetwork model. In some implementations, we can first determine all hyperedges including the second node, and then determine all nodes within these hyperedges except the second node as adjacent nodes of the second node. Figure 3 For example, assuming v6 is the node corresponding to a trending topic, then the adjacent nodes of v6 include v1 to v5, and v7. In other implementations, nodes directly connected to the second node can be defined as the adjacent nodes of the second node. Figure 3For example, if v6 is the node corresponding to a popular topic, then the adjacent nodes of v6 include v1, v3, v5 and v7.

[0123] For each adjacent node, we can query the number of superedges in the supernetwork model of topic T that simultaneously include the second node and that adjacent node. Then, we can select the topics corresponding to the first N adjacent nodes in descending order of these numbers to determine the popular related topics of the second topic.

[0124] In this embodiment, after identifying popular topics, related popular topics can be further identified based on these popular topics, thereby expanding the scope of topic monitoring and further improving the reliability of network monitoring.

[0125] It should be noted that the various optional implementation methods described in the embodiments of this application can be combined with each other or implemented individually without conflict, and the embodiments of this application do not limit this.

[0126] For ease of understanding, the following example is provided:

[0127] In some examples, network monitoring methods can be as follows: Figure 4 As shown:

[0128] exist Figure 4 In China, online platforms manifest as communities, and topic sources manifest as posts.

[0129] Step 401: Initialize the community post set for time period T.

[0130] In practice, all posts and comments are segmented into words, and punctuation and stop words are removed to obtain a preliminary set of topics and posts.

[0131] Step 402: Construct the topic supernetwork model for time period T.

[0132] In practice, a topic hypernetwork model is constructed based on the hierarchical relationship between topics and posts.

[0133] Step 403: Read the unoperated node v(T).

[0134] Unoperated nodes v(T) are nodes corresponding to unclassified topics. In some embodiments, unoperated nodes and operated nodes can be distinguished in the topic hypernetwork model, such as by labeling, which can improve the reading efficiency of unoperated nodes.

[0135] In practice, the nodes in the topic supernetwork are traversed to determine if there are any nodes that have not been operated on. If so, the process proceeds to step 404; otherwise, the operation ends.

[0136] Step 404, determine q v(T) Whether it is greater than the threshold.

[0137] q v (T) represents the node weight of v(T) in the topic supernetwork during time period T, and this threshold is the weight threshold. If it is greater than the threshold, proceed to step 408; if it is less than or equal to the threshold, proceed to step 405.

[0138] Step 405, determine q v Is (T-1) greater than the threshold?

[0139] q v (T-1) is the node weight of v(T) during the time period T-1. If yes, proceed to step 406; otherwise, proceed to step 407.

[0140] Step 406: Determine that the topic corresponding to node v(T) is a decay topic.

[0141] Step 407: Determine that the topic corresponding to node v(T) is a non-popular topic.

[0142] Step 408: Calculate the betweenness centrality and node superdegree of v(T).

[0143] Step 409: Determine whether the betweenness centrality and node overdegree of v(T) are greater than the corresponding thresholds.

[0144] If not, proceed to step 410; if yes, proceed to step 411.

[0145] Step 410, determine q v (T)-q v Is (T-1) greater than the threshold?

[0146] If yes, proceed to step 411; otherwise, proceed to step 407.

[0147] Step 411: Determine that the topic corresponding to node v(T) is a popular topic.

[0148] Step 412: The topics corresponding to the top N nodes with the largest number of shared superedges with node v(T) are identified as popular related topics.

[0149] Step 413: Determine if there are any unoperated nodes.

[0150] If yes, return to step 403; otherwise, end the calculation.

[0151] As can be seen, this application proposes a topic-based hypernetwork model, using topics as nodes and topic sources as hyperedges to describe the relationships between topics and topic sources. By incorporating the time dimension into the hypernetwork model, a topic hypernetwork evolution model is constructed to describe the changes in topics over time. Based on the hypernetwork dynamic evolution model, the weight of topic nodes, betweenness centrality, and hyperdegree—three network topology indicators—are used as classification rules for topics. Topics can be divided into four categories: popular topics, unpopular topics, declining topics, and popular related topics, enabling comprehensive monitoring of online public opinion.

[0152] Based on the network monitoring method provided in the above embodiments, this application also provides specific implementations of a network monitoring device. Please refer to the following embodiments.

[0153] See Figure 5 The network monitoring device provided in this application embodiment may include:

[0154] The acquisition module 501 is used to acquire information related to the publication of a network platform in T time periods. The information related to the publication of the network platform in the t-th time period includes: r topic sources included in the network platform in the t-th time period, f topics published through the r topic sources in the t-th time period, and the publication frequency of each of the f topics in the t-th time period; T is a positive integer, t is a positive integer less than or equal to T, r is a positive integer, and f is an integer greater than 1.

[0155] Module 502 is used to construct T topic hypernetwork models corresponding one-to-one with the T time periods using the release-related information of the T time periods. The t-th topic hypernetwork model corresponding to the t-th time period includes a node set, a node weight set, and a hyperedge set. The node set includes f nodes corresponding one-to-one with the f topics. The node weight set includes f node weights corresponding one-to-one with the f nodes, and the node weight of each node is determined based on the release frequency of the topic corresponding to the node within the t-th time period. The hyperedge set includes r hyperedges corresponding one-to-one with the r topic sources, and the hyperedge corresponding to each topic source includes n nodes corresponding one-to-one with the n topics released through the topic source; n is a positive integer less than f.

[0156] The classification module 503 is used to classify topics in the network platform using the T topic hypernetwork models.

[0157] In some embodiments, the classification module includes:

[0158] The first acquisition submodule is used to obtain the weight of the Tth node corresponding to the first node from the Tth topic hypernetwork model corresponding to the Tth time period;

[0159] The first determining submodule is used to determine the betweenness centrality and node superdegree of the first node according to the topic supernetwork model when the weight of the T-th node is greater than the weight threshold. The betweenness centrality is used to characterize the contribution of the first topic corresponding to the first node to other topics. The node superdegree is used to characterize the number of hyperedges including the first node in the topic supernetwork model.

[0160] The second determining submodule is used to determine the first topic as a popular topic when the intermediary centrality is greater than the centrality threshold and the node extremum is greater than the extremum threshold.

[0161] In some embodiments, the first determining submodule includes:

[0162] A connection unit is used to fully connect the nodes in each hyperedge of the T-th topic hypernetwork model to obtain the fully connected graph corresponding to the T-th topic hypernetwork model.

[0163] The query unit is used to query the fully connected graph to obtain the contribution of the first node to each node group, wherein each node group includes any two nodes in the T-th topic hypernetwork model other than the first node, and at least one node in the different nodes is different; the contribution of the first node to each node group is determined based on the shortest connection path between the two nodes in the node group in the fully connected graph.

[0164] The determining unit is used to determine the betweenness centrality of the first node based on the contribution of the first node to each node group.

[0165] In some embodiments, the query unit is specifically used for:

[0166] For each node group, query the fully connected graph to obtain g shortest connection paths between two nodes in the node group in the fully connected graph, and h shortest connection paths including the first node, where g is a positive integer and h is a positive integer less than or equal to g.

[0167] The ratio of h to g is determined as the contribution of the first node to the node group.

[0168] In some embodiments, the classification module further includes:

[0169] The second acquisition submodule is used to acquire the weight of the first node corresponding to the first node from the super network model of the topic corresponding to the time period T-1.

[0170] The third determining submodule is used to determine that the first topic is a non-popular topic when both the weight of the T-th node and the weight of the (T-1)-th node are less than or equal to the weight threshold; or...

[0171] The fourth determining submodule is used to determine the first topic as a popular topic if, when the intermediary centrality is less than or equal to the centrality threshold, or the node exponentiation is less than the exponentiation threshold, the difference between the weight of the T-th node and the weight of the (T-1)-th node is greater than a difference threshold; or...

[0172] The fifth determining submodule is used to determine the first topic as a declining topic when the weight of the T-th node is less than or equal to the weight threshold, but the weight of the (T-1)-th node is greater than the weight threshold.

[0173] In some embodiments, the classification module includes:

[0174] The sixth determination submodule is used to determine at least two neighboring nodes of the second node in the supernetwork model of the Tth topic corresponding to the Tth time period, wherein the second topic corresponding to the second node is a hot topic;

[0175] The third acquisition submodule is used to acquire the number of shared super edges between each of the at least two adjacent nodes and the second node, wherein the shared super edges are the super edges in the T-th topic hypernetwork model that include the adjacent node and the second node.

[0176] The seventh determination submodule is used to determine the third topic corresponding to the third node as a popular related topic, wherein the third node is the Nth neighboring node among the at least two neighboring nodes that have the most shared superedges with the second node, and N is a positive integer.

[0177] In some embodiments, the information to be published at the t-th time period further includes: the user attention of each topic source among the r topic sources during the t-th time period;

[0178] The t-th topic hypernetwork model also includes a set of advanced weights, wherein the set of hyperedge weights includes r hyperedge weights that correspond one-to-one with the r hyperedges, and the hyperedge weight corresponding to each hyperedge is determined based on the user attention of the hyperedge in the t-th time period.

[0179] The network monitoring device provided in this application embodiment can implement the various processes in the method embodiment, and will not be described again here to avoid repetition.

[0180] Figure 6 A schematic diagram of the hardware structure for network monitoring provided in an embodiment of this application is shown.

[0181] The network monitoring device may include a processor 601 and a memory 602 storing computer program instructions.

[0182] Specifically, the processor 601 may include a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits that can be configured to implement the embodiments of this application.

[0183] Memory 602 may include mass storage for data or instructions. For example, and not limitingly, memory 602 may include a hard disk drive (HDD), floppy disk drive, flash memory, optical disk, magneto-optical disk, magnetic tape, or Universal Serial Bus (USB) drive, or a combination of two or more of these. Where appropriate, memory 602 may include removable or non-removable (or fixed) media. Where appropriate, memory 602 may be internal or external to the integrated gateway disaster recovery device. In a particular embodiment, memory 602 is non-volatile solid-state memory.

[0184] Memory may include read-only memory (ROM), random access memory (RAM), disk storage media devices, optical storage media devices, flash memory devices, and electrical, optical, or other physical / tangible memory storage devices. Therefore, typically, memory includes one or more tangible (non-transitory) computer-readable storage media (e.g., memory devices) encoded with software including computer-executable instructions, and when the software is executed (e.g., by one or more processors), it is operable to perform the operations described with reference to the method according to one aspect of this disclosure.

[0185] The processor 601 implements any of the network monitoring methods described in the above embodiments by reading and executing computer program instructions stored in the memory 602.

[0186] In one example, the network monitoring device may also include a communication interface 606 and a bus 610. Wherein, for example... Figure 6 As shown, the processor 601, memory 602, and communication interface 606 are connected through bus 610 and complete communication with each other.

[0187] The communication interface 606 is mainly used to realize communication between various modules, devices, units and / or equipment in the embodiments of this application.

[0188] Bus 610 includes hardware, software, or both, that couples components of a network monitoring device together. For example, and not limitingly, the bus may include an Accelerated Graphics Port (AGP) or other graphics bus, an Enhanced Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), HyperTransport (HT) interconnect, an Industry Standard Architecture (ISA) bus, an Infinite Bandwidth Interconnect, a Low Pin Count (LPC) bus, a memory bus, a Microchannel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a Video Electronics Standards Association Local (VLB) bus, or other suitable buses, or combinations of two or more of these. Where appropriate, bus 610 may include one or more buses. Although specific buses are described and illustrated in embodiments of this application, any suitable bus or interconnect is contemplated herein.

[0189] Furthermore, in conjunction with the network monitoring methods described in the above embodiments, this application embodiment can provide a computer storage medium for implementation. This computer storage medium stores computer program instructions; when these computer program instructions are executed by a processor, they implement any of the network monitoring methods described in the above embodiments.

[0190] This application also provides a computer program, including computer-readable code, which, when executed by a processor, implements any of the network monitoring methods described above.

[0191] This application provides a computer program product in which the instructions are executed by the processor of an electronic device, causing the electronic device to perform any of the network monitoring methods described in the above embodiments.

[0192] It should be clarified that this application is not limited to the specific configurations and processes described above and shown in the figures. For the sake of brevity, detailed descriptions of known methods are omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of this application is not limited to the specific steps described and shown. Those skilled in the art can make various changes, modifications, and additions, or change the order of steps, after understanding the spirit of this application.

[0193] The functional blocks shown in the above-described structural diagram can be implemented as hardware, software, firmware, or a combination thereof. When implemented in hardware, they can be, for example, electronic circuits, application-specific integrated circuits (ASICs), appropriate firmware, plug-ins, function cards, etc. When implemented in software, the elements of this application are programs or code segments used to perform the required tasks. Programs or code segments can be stored on a machine-readable medium or transmitted over a transmission medium or communication link via data signals carried on a carrier wave. "Machine-readable medium" can include any medium capable of storing or transmitting information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, ROM, flash memory, erasable ROM (EROM), floppy disks, CD-ROMs, optical disks, hard disks, fiber optic media, radio frequency (RF) links, etc. Code segments can be downloaded via computer networks such as the Internet, intranets, etc.

[0194] It should also be noted that the exemplary embodiments mentioned in this application describe methods or systems based on a series of steps or apparatus. However, this application is not limited to the order of the above steps; that is, the steps can be performed in the order mentioned in the embodiments, or in a different order, or several steps can be performed simultaneously.

[0195] The aspects of this disclosure have been described above with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this disclosure. It should be understood that each block in 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, a special-purpose computer, or other programmable data processing apparatus to produce a machine such that these instructions, executable via the processor of the computer or other programmable data processing apparatus, enable the implementation of the functions / actions specified in one or more blocks of the flowchart illustrations and / or block diagrams. Such a processor can be, but is not limited to, a general-purpose processor, a special-purpose processor, a special application processor, or a field-programmable logic circuit. It is also understood that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can also be implemented by special-purpose hardware performing the specified functions or actions, or can be implemented by a combination of special-purpose hardware and computer instructions.

[0196] The above description is merely a specific implementation of this application. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, modules, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here. It should be understood that the protection scope of this application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and these modifications or substitutions should all be covered within the protection scope of this application.

Claims

1. A network monitoring method, characterized in that, include: Obtain relevant information published by a network platform over T time periods, wherein the relevant information published by the network platform in time period t includes: r topic sources included in the network platform in time period t, f topics published through the r topic sources in time period t, and the publication frequency of each of the f topics in time period t; T is a positive integer, t is a positive integer less than or equal to T, r is a positive integer, and f is an integer greater than 1; Using the publication information from the T time periods, T topic hypernetwork models corresponding one-to-one with the T time periods are constructed. The t-th topic hypernetwork model corresponding to the t-th time period includes a node set, a node weight set, and a hyperedge set. The node set includes f nodes corresponding one-to-one with the f topics. The node weight set includes f node weights corresponding one-to-one with the f nodes, and the node weight of each node is determined based on the publication frequency of the topic corresponding to the node within the t-th time period. The hyperedge set includes r hyperedges corresponding one-to-one with the r topic sources, and the hyperedge corresponding to each topic source includes n nodes corresponding one-to-one with the n topics published through the topic source; n is a positive integer less than f. The topics in the network platform are classified using the T topic hypernetwork models.

2. The method according to claim 1, characterized in that, The process of classifying topics on the network platform using the T topic hypernetwork models includes: From the supernetwork model of topic T corresponding to time period T, obtain the weight of node T corresponding to the first node; When the weight of the T-th node is greater than the weight threshold, the betweenness centrality and node superdegree corresponding to the first node are determined according to the T-th topic hypernetwork model. The betweenness centrality is used to characterize the contribution of the first topic corresponding to the first node to other topics; the node superdegree is used to characterize the number of hyperedges including the first node in the T-th topic hypernetwork model. If the centrality of the intermediaries is greater than the centrality threshold and the exponent degree of the nodes is greater than the exponent degree threshold, the first topic is determined to be a popular topic.

3. The method according to claim 2, characterized in that, The step of determining the betweenness centrality and node superdegree of the first node according to the T-th topic hypernetwork model includes: By fully connecting the nodes in each hyperedge of the T-th topic hypernetwork model, the fully connected graph corresponding to the T-th topic hypernetwork model is obtained. The contribution of the first node to each node group is obtained by querying the fully connected graph. Each node group includes any two nodes in the T-th topic hypernetwork model other than the first node, and at least one node in each node group is different. The contribution of the first node to each node group is determined based on the shortest connection path between the two nodes in the node group in the fully connected graph. The intermediary centrality of the first node is determined based on its contribution to each node group.

4. The method according to claim 3, characterized in that, The querying of the fully connected graph to obtain the contribution of the first node to each node group includes: For each node group, query the fully connected graph to obtain g shortest connection paths between two nodes in the node group in the fully connected graph, and h shortest connection paths including the first node, where g is a positive integer and h is a positive integer less than or equal to g. The ratio of h to g is determined as the contribution of the first node to the node group.

5. The method according to claim 2, characterized in that, After obtaining the weight of the Tth node corresponding to the first node from the Tth topic hypernetwork model corresponding to the Tth time period, the method further includes: Obtain the weight of the first node corresponding to the first node from the super network model of the topic corresponding to the time period T-1; If both the weight of node T and the weight of node T-1 are less than or equal to the weight threshold, the first topic is determined to be a non-popular topic; or, If the centrality of a node is less than or equal to the centrality threshold, or the node exponentiation is less than the exponentiation threshold, and the difference between the weight of the T-th node and the weight of the (T-1)-th node is greater than the difference threshold, then the first topic is determined to be a popular topic; or... If the weight of node T is less than or equal to the weight threshold, but the weight of node T-1 is greater than the weight threshold, the first topic is determined to be a declining topic.

6. The method according to any one of claims 1 to 5, characterized in that, The process of classifying topics on the network platform using the T topic hypernetwork models includes: Determine at least two neighboring nodes of the second node in the supernetwork model of the Tth topic corresponding to the Tth time period, wherein the second topic corresponding to the second node is a hot topic; For each of the at least two adjacent nodes, obtain the number of shared super edges between the adjacent node and the second node, wherein the shared super edges are the super edges in the T-th topic supernetwork model that include the adjacent node and the second node; The third topic corresponding to the third node is determined as a popular related topic. The third node is the Nth neighboring node among the at least two neighboring nodes that has the most shared superedges with the second node, where N is a positive integer.

7. The method according to claim 1, characterized in that, The information released at time t also includes: the user attention of each topic source among the r topic sources in the time period t; The t-th topic hypernetwork model also includes a set of advanced weights, wherein the set of hyperedge weights includes r hyperedge weights that correspond one-to-one with the r hyperedges, and the hyperedge weight corresponding to each hyperedge is determined based on the user attention of the hyperedge in the t-th time period.

8. A network monitoring device, characterized in that, The device includes: The acquisition module is used to acquire relevant information published by the network platform in T time periods. The relevant information published by the network platform in the t-th time period includes: r topic sources included in the network platform in the t-th time period, f topics published through the r topic sources in the t-th time period, and the publication frequency of each of the f topics in the t-th time period; T is a positive integer, t is a positive integer less than or equal to T, r is a positive integer, and f is an integer greater than 1. A construction module is used to construct T topic hypernetwork models corresponding one-to-one with the T time periods using the release-related information of the T time periods. The t-th topic hypernetwork model corresponding to the t-th time period includes a node set, a node weight set, and a hyperedge set. The node set includes f nodes corresponding one-to-one with the f topics. The node weight set includes f node weights corresponding one-to-one with the f nodes, and the node weight of each node is determined based on the release frequency of the topic corresponding to the node within the t-th time period. The hyperedge set includes r hyperedges corresponding one-to-one with the r topic sources, and the hyperedge corresponding to each topic source includes n nodes corresponding one-to-one with the n topics released through the topic source; n is a positive integer less than f. The classification module is used to classify topics in the network platform using the T topic hypernetwork models.

9. A network monitoring device, characterized in that, The device includes: a processor and a memory storing computer program instructions; the processor, when executing the computer program instructions, implements the network monitoring method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer program instructions that, when executed by a processor, implement the network monitoring method as described in any one of claims 1 to 6.

11. A computer program comprising computer-readable code, characterized in that, When the computer-readable code is executed by a processor, it implements the network monitoring method as described in any one of claims 1 to 7.

12. A computer program product, characterized in that, When the instructions in the computer program product are executed by the processor of the electronic device, the electronic device performs the network monitoring method as described in any one of claims 1 to 7.

Citation Information

Patent Citations

  • News theme mining method based on multiple views

    CN115203313A

  • Network hot topic propagation mode classification method and system integrating multiple models

    CN116467454A