A social cache retrieval system with fusion hotspot analysis

Abstract

The application relates to the technical field of data caching, in particular to a social caching retrieval system fusing hot spot analysis, which comprises a hot degree modeling unit, a social association unit and a caching resource allocation unit. The hot degree modeling unit is used for obtaining a hot spot event set, performing hot degree modeling on each hot spot event in the hot spot event set, and obtaining hot degree characteristics of the hot spot event. The social association unit is used for obtaining a user set and constructing a user-content bipartite graph according to interactive behaviors between the hot spot event set and the user set. Through the hot degree modeling unit and the propagation calculation unit, the hot spot event can be modeled according to user behaviors and historical data, and the hot degree of the hot spot event at a future moment can be predicted by using a hot degree prediction model, so that the events which will become hot spots can be identified in advance, and more caching resources can be allocated to the events according to the prediction results.

Description

A social caching retrieval system integrating hotspot analysis Technical Field

[0001] This invention relates to the field of data caching technology, and more specifically to a social caching retrieval system that integrates hotspot analysis. Background Technology

[0002] Traditional systems typically lack prediction and popularity analysis for trending events, making it impossible to identify events poised to become trending. Consequently, when a trending event suddenly emerges, traditional systems may fail to load it into the cache in time, leading to delays or cache misses for user requests. Furthermore, traditional systems often use fixed strategies for data storage, such as first-come, first-served caching or random data deletion when cache capacity is limited. This approach doesn't consider the popularity and social relevance of events, failing to optimize caching for trending events and potentially resulting in slow response times for popular content. Moreover, traditional systems usually manage caches only based on the most frequently accessed data, ignoring the impact of social networks and user behavior on content demand. For example, without adjusting cached content based on user social relationships and behavioral analysis, the system cannot determine which events will generate widespread attention on social networks, leading to low cache hit rates. Finally, because traditional systems lack fine-grained management and prioritization of cached content, many unpopular or irrelevant events still occupy cache space, reducing cache space utilization efficiency, under-optimizing resources, and resulting in poor performance. Summary of the Invention

[0003] The technical problem to be solved by the present invention is to overcome the shortcomings of the prior art and provide a social cache retrieval system that integrates hotspot analysis.

[0004] The technical solution adopted to solve the above-mentioned technical problems is: a social caching retrieval system integrating hotspot analysis, comprising:

[0005] A heat modeling unit is used to acquire a set of hot events and perform heat modeling on each hot event in the set of hot events to obtain the heat characteristics of the hot events.

[0006] A social association unit is used to obtain a user set, construct a user-content bipartite graph based on the interaction behavior between the hot event set and the user set, and determine the global social association degree of each hot event in the hot event set based on the user-content bipartite graph.

[0007] The propagation calculation unit is used to predict the popularity characteristics of the hot topic event based on a pre-trained popularity prediction model to obtain the future popularity characteristics of the hot topic event, and to propagate the content based on the future popularity characteristics of the hot topic event according to an infectious disease model to obtain the diffusion probability of the hot topic event.

[0008] A cache calculation unit is configured to define a cache utility function based on the global social relevance, future popularity characteristics, and diffusion probability of the hot events, and to determine the cache utility value of each hot event in the set of hot events based on the cache utility function.

[0009] Preferably, the system further includes:

[0010] A social caching unit is configured to determine a list of hot events and a list of non-hot events based on the caching utility value of the hot events, store the list of hot events in a local cache, and store the list of non-hot events in each cache node of the cache cluster.

[0011] The retrieval request unit is used to obtain a retrieval request, match the retrieval request with the hot events in the local cache, and if no match is found, send a query request to the cache node with the highest cache utility value within a preset number of hops according to the distributed hash table.

[0012] Preferably, a heat model is performed on each of the hot events in the hot event set to obtain the heat characteristics of the hot events, including:

[0013] The number of visits to the hot events is counted according to a preset time sliding window.

[0014] The total number of reposts of hot events in the user set is obtained according to a preset time sliding window, and the social influence weight of the user is determined according to the total number of reposts of hot events in the user set.

[0015] The user's interactive behavior within the time sliding window is obtained, and the intensity of the user's interactive behavior within the time sliding window is defined based on the user's interactive behavior within the time sliding window. The interactive behavior includes liking, commenting, and forwarding, and the intensity of the interactive behavior is a weighted sum of the interactive behaviors.

[0016] The popularity characteristics of a hot topic are determined based on the number of times the hot topic is accessed, the social influence weight of the users who interact with the hot topic, and the intensity of the interactive behavior of the users who interact with the hot topic.

[0017] Preferably, the formula for calculating the user's social influence weight is as follows:

[0018]

[0019] in, Represents the k-th user u in the user set U. k For the i-th hot event c in the set of hot events i Social influence weight, SC(u k ,c i ) represents the k-th user u in the user set. k For the i-th hot event c in the set of hot events i The number of forwards, SC(U,c i ) represents the user set U and the i-th hot event c in the hot event set. i Total number of reposts;

[0020] The formula for calculating the popularity characteristics of the trending events is as follows:

[0021]

[0022] Among them, H i (t) represents the popularity characteristics of hot events within a sliding time window t, λ represents the adjustable optimization weight, and N access (c i (t) represents the user's response to the i-th hot event c in the set of hot events. i Number of visits to trending events I(u) represents the total number of times a user accesses the hot events in the hot event set. k ,c i (t) represents the user's view of the i-th hot event c in the hot event set within the time sliding window t. i The intensity of interactive behavior, This represents the function for finding the maximum value.

[0023] Preferably, the user-content bipartite graph includes a set of user nodes, a set of hot event nodes, a set of connecting edges, and a set of edge weights. The user nodes correspond to users, the hot event nodes correspond to hot events, and the connecting edges are used to connect the user nodes and the hot event nodes. When a user interacts with a hot event, there is a connecting edge between the corresponding user node and the hot event node. The edge weights of the connecting edges are defined according to the intensity of the interaction between the user and the hot event.

[0024] Preferably, determining the global social relevance of each of the hot topics in the hot topic event set based on the user-content bipartite graph includes:

[0025] An initial influence value is assigned to each node in the user-content bipartite graph, wherein the formula for calculating the initial influence value is as follows:

[0026]

[0027] Among them, PR (0) (k) represents the initial influence value of the k-th node in the user-content bipartite graph, and N represents the number of nodes in the user-content bipartite graph;

[0028] The steady-state influence of each node in the user-content bipartite graph is determined using random walk simulation, and the formula for calculating the steady-state influence is as follows:

[0029]

[0030] Among them, PR (l+1) (k) represents the steady-state influence of a node in the (l+1)th iteration, d represents the damping factor, and the damping factor represents the probability that the user continues the random walk. The remainder (1-d) is the probability of randomly jumping to any node. B(k) represents the set of source nodes of the node, and M bk Let represent the transition probability from node b to k, and A bk Let $\mathbf$ represent the edge weight from node $b$ to node $k$ in the user-content bipartite graph.

[0031] Repeat the above operation until the steady-state influence after two iterations is less than the preset influence threshold, so as to obtain the final influence threshold.

[0032] The final steady-state influence of a node is normalized based on the maximum and minimum values ​​of the final steady-state influence of all nodes to obtain the global social correlation of the node, which is the global social correlation of the hot topic event.

[0033] Preferably, the heat prediction model uses an LSTM network, and the formula for calculating the diffusion probability of the hot event is as follows:

[0034]

[0035] Among them, P spread (c i ) represents the diffusion probability of a hot topic event, β represents the preset density coefficient, γ represents the preset decay factor, and S i This indicates the overall social relevance of trending events. This indicates the future popularity characteristics.

[0036] Preferably, the formula for calculating the caching utility value of the hot events is as follows:

[0037]

[0038] Among them, U i This represents the caching utility value of hot events, α represents the preset weighting factor, and Size(c i ) indicates the data size of hot events.

[0039] The beneficial effects of the present invention are as follows: (1) The present invention, through the heat modeling unit and the propagation calculation unit, can model the heat of hot events based on user behavior and historical data, and use the heat prediction model to predict the heat of hot events in the future, thereby identifying events that are about to become hot events in advance, and allocating more cache resources to these events based on the prediction results. In addition, through the social association unit, a user-content bipartite graph is constructed based on the interaction between users and hot events, and the global social association degree is calculated, so that the system can identify events that are not only hot in time, but also highly associated in social networks and may attract a large number of user attention, thereby more accurately allocating the cache priority of hot events; (2) The retrieval request unit of the present invention can quickly match with the local cached list of hot events, reduce query delay, and quickly return results for stored hot events, improving the user's retrieval experience. In addition, when no match is found in the local cache, the system sends a query request to the cache node with the highest cache utility value in the cache cluster through a distributed hash table to ensure that the remote cache is used for the query. The speed and efficiency of the storage node to obtain hot events avoid the bottleneck of frequently obtaining data from the database or other slow storage; (3) The present invention defines the cache utility function based on the global social correlation, future hotness characteristics and diffusion probability of hot events through the cache computing unit, which can accurately evaluate the cache value of each hot event. This process can ensure that the hot events most likely to be retrieved are stored in the local cache first, improving the cache utilization and hit rate. Furthermore, the social cache unit stores hot events and non-hot events separately according to the cache utility value. Hot events are stored in the local cache, while non-hot events are stored in the cluster cache. This can reduce the waste of cache space and ensure that hot events can quickly respond to user requests. By storing the list of hot events in the local cache and storing non-hot events in different nodes of the cache cluster, this distributed caching method makes the system have good scalability. As hot events change, the system can flexibly adjust the caching strategy and distribute the load among multiple cache nodes as needed, thereby adapting to changes in user needs. Attached Figure Description

[0040] Figure 1 is a schematic diagram of the overall system architecture in one embodiment of the present invention.

[0041] Figure labels: 1. Popularity modeling unit; 2. Social association unit; 3. Propagation calculation unit; 4. Cache calculation unit; 5. Social caching unit; 6. Retrieval request unit. Detailed Implementation

[0042] Example 1, as shown in Figure 1, presents a social caching retrieval system integrating hotspot analysis, comprising:

[0043] Popularity Modeling Unit 1 is used to obtain a set of popular events and perform popularity modeling on each popular event in the set of popular events to obtain the popularity characteristics of popular events.

[0044] Social Relation Unit 2 is used to obtain the user set, construct a user-content bipartite graph based on the interaction behavior between the hot event set and the user set, and determine the global social relevance of each hot event in the hot event set based on the user-content bipartite graph;

[0045] The propagation calculation unit 3 is used to predict the popularity characteristics of hot events based on a pre-trained popularity prediction model to obtain the future popularity characteristics of hot events, and to propagate the content based on the future popularity characteristics of hot events according to an infectious disease model to obtain the diffusion probability of hot events.

[0046] The cache calculation unit 4 is used to define a cache utility function based on the global social relevance of hot events, the future popularity characteristics and diffusion probability of hot events, and to determine the cache utility value of each hot event in the hot event set based on the cache utility function.

[0047] In this invention, popularity refers to the degree of attention a certain event, content, or topic receives, which is usually measured by user interaction (such as clicks, comments, and shares). Popularity modeling is used to predict the degree of attention an event or content receives by analyzing historical data and user behavior. A set of hot events refers to a set of events or topics that are currently being widely discussed and followed on social platforms. A bipartite graph is a graph structure in which nodes can be divided into two categories: users and content. On social media platforms, there is a relationship between users and content, such as a user's likes or comments on content. The edges of a bipartite graph connect users to the content they interact with (trending events). A trending topic prediction model is a machine learning or statistical model designed to predict the future popularity of trending topics using historical data. For example, it can predict the change in the popularity of an event over a future period by using users' historical behavior data and current popularity characteristics. Infectious disease models are inspired by infectious disease transmission models in epidemiology, such as the SIR model (Susceptible, Infected, Recovered). In social media, the spread of trending topics is similar to the spread of infectious diseases. Users spread content to other users through interactions (such as forwarding and commenting). This model calculates the diffusion probability of a trending topic by simulating this diffusion process, that is, to what extent the event will receive more attention from users in the future. Diffusion probability refers to the probability that a trending topic will receive more attention from users at a certain point in the future.

[0048] In an optional embodiment, the system further includes:

[0049] Social caching unit 5 is used to determine the list of hot events and the list of non-hot events based on the caching utility value of hot events, store the list of hot events in the local cache, and store the list of non-hot events in each cache node in the cache cluster;

[0050] The retrieval request unit 6 is used to obtain retrieval requests and match them with hot events in the local cache. If no match is found, a query request is sent to the cache node with the highest cache utility value within a preset number of hops according to the distributed hash table.

[0051] It's important to note that cache utility value is a metric used to measure whether a trending event should be cached. It's typically based on factors such as the event's popularity, probability of spread, and user engagement. Events with high utility values ​​are more likely to be frequently accessed in the future and therefore more worthy of caching. A trending event list refers to a collection of events with high cache utility values. Because these events have received significant attention and are likely to continue to be queried or discussed by users, they are stored in a local cache for fast access. Non-trending events have lower cache utility values, meaning they have lower popularity or a lower probability of future access. Therefore, they are not stored in a local cache but are distributed across a cache cluster to reduce the burden on the local cache. A local cache refers to a cache stored on a single server or device, used to quickly respond to user requests. Trending events, due to their high access frequency, are prioritized for storage in the local cache, thereby improving retrieval efficiency. A cache cluster (Cache Cluster)... A cluster is a distributed caching system composed of multiple cache nodes. Non-hot events are stored in this cluster, which reduces the burden on individual cache nodes and ensures system scalability and efficiency. A retrieval request refers to a user's query request, where the user wants to retrieve certain information or events. In this system, a retrieval request is a query for information about hot events. Hot event matching in the local cache means that when a user issues a query, the retrieval request unit first searches the local cache for matching hot events. If the query is successful, the system directly returns the cached data, thereby improving response speed. Distributed hash tables (Distributed Hash Tables) are also used. A Distributed Hash Table (DHT) is a distributed storage system that distributes data across multiple nodes and provides efficient lookup and access mechanisms. In this system, DHT is used to distribute and manage cached data, using hash values ​​to locate storage nodes. The preset hop count refers to the maximum number of nodes the system allows for a query when using the distributed hash table. A smaller hop count results in a faster response time, while a larger hop count may lead to a longer query time. The preset hop count is a parameter used to limit the query range and ensure query efficiency. The cache node with the highest cache utility value refers to the node with the highest utility value that the retrieval request unit will send a request to other cache nodes based on the distributed hash table if there is no matching hot event in the local cache. This node increases the probability of a successful retrieval because the events stored on these nodes are more likely to be currently popular, thus increasing the chance of a successful match.

[0052] Example 2: The social caching retrieval system integrating hotspot analysis proposed in this invention, compared with Example 1, further includes: performing heat modeling on each hotspot event in the hotspot event set to obtain the heat features of the hotspot events, including:

[0053] The number of visits to trending events is counted based on a preset time sliding window.

[0054] The total number of reposts of trending events for a user set is obtained based on a preset time sliding window, and the user's social influence weight is determined based on the total number of reposts of trending events for the user set.

[0055] Acquire user interaction behavior within a time-sliding window, and define the intensity of user interaction behavior within the time-sliding window based on the user's interaction behavior within the time-sliding window. The interaction behavior includes likes, comments, and shares, and the intensity of the interaction behavior is the weighted sum of the interaction behaviors.

[0056] The popularity characteristics of trending events are determined based on the number of times a trending event is accessed, the social influence weight of users interacting with the trending event, and the intensity of the interactive behavior of users interacting with the trending event.

[0057] In this embodiment, the time sliding window is a dynamic time range used to collect and analyze data within a specific time period. The size and sliding step of the time window are preset. As time progresses, the window continuously moves forward, thus including new data and excluding old data. Sliding windows are widely used to process and analyze data with time-series characteristics, such as the number of visits to trending events and user behavior data. The number of forwards reflects the extent to which users spread an event; the higher the number of forwards, the wider the spread of the event, and the greater the potential impact on other users.

[0058] In an optional embodiment, the formula for calculating a user's social influence weight is as follows:

[0059]

[0060] in, Represents the k-th user u in the user set U. k For the i-th hot event c in the set of hot events i Social influence weight, SC(u k ,c i ) represents the k-th user u in the user set. k For the i-th hot event c in the set of hot events i The number of forwards, SC(U,c i ) represents the user set U and the i-th hot event c in the hot event set. i Total number of reposts;

[0061] The formula for calculating the popularity characteristics of trending events is as follows:

[0062]

[0063] Among them, H i (t) represents the popularity characteristics of hot events within a sliding time window t, λ represents the adjustable optimization weight, and N access (c i (t) represents the user's response to the i-th hot event c in the set of hot events. i Number of visits to trending events I(u) represents the total number of times a user accesses the hot events in the hot event set. k ,c i (t) represents the user's view of the i-th hot event c in the hot event set within the time sliding window t. i The intensity of interactive behavior, This represents the function for finding the maximum value.

[0064] In an optional embodiment, the user-content bipartite graph includes a set of user nodes, a set of hot event nodes, a set of connecting edges, and a set of edge weights. User nodes correspond to users, hot event nodes correspond to hot events, and connecting edges are used to connect user nodes and hot event nodes. When a user interacts with a hot event, there is a connecting edge between the corresponding user node and the hot event node. The edge weights of the connecting edges are defined according to the intensity of the interaction between the user and the hot event.

[0065] It should be noted that in a user-content bipartite graph, the graph structure consists of two types of nodes (user nodes and trending event nodes) and the edges connecting them. The edges represent the interaction between users and trending events, and the edge weights reflect the intensity of the interaction. The intensity of the interaction serves as the source of the edge weights, and the weights of the edges are defined by the user's behavior with the event (such as liking, commenting, and forwarding), thereby revealing the degree of user attention and participation in the event. This graph model is commonly used in social network analysis, user behavior prediction, and research on the spread of trending events.

[0066] In an optional embodiment, determining the global social relevance of each hot event in the hot event set based on the user-content bipartite graph includes:

[0067] Assign initial influence values ​​to each node in the user-content bipartite graph. The formula for calculating the initial influence value is as follows:

[0068]

[0069] Among them, PR(0) (k) represents the initial influence value of the k-th node in the user-content bipartite graph, and N represents the number of nodes in the user-content bipartite graph;

[0070] The steady-state influence of each node in the user-content bipartite graph is determined using random walk simulations. The formula for calculating the steady-state influence is as follows:

[0071]

[0072] Among them, PR (l+1) (k) represents the steady-state influence of a node in the (l+1)th iteration, d represents the damping factor, and the damping factor represents the probability that the user continues the random walk. The remainder (1-d) is the probability of randomly jumping to any node. B(k) represents the set of source nodes of the node, and M bk Let represent the transition probability from node b to k, and A bk Let $\mathbf$ represent the edge weight from node $b$ to node $k$ in the user-content bipartite graph.

[0073] Repeat the above operation until the steady-state influence after two iterations is less than the preset influence threshold, so as to obtain the final influence threshold.

[0074] The final steady-state influence of a node is normalized based on the maximum and minimum values ​​of the final steady-state influence of all nodes to obtain the global social correlation of the node, which is the global social correlation of the hot topic event.

[0075] It should be noted that the initial influence value is the influence distribution of each node (whether a user node or a content node) before any iterations; steady-state influence refers to the influence value of a node in a stable state after multiple iterations. When the random walk model is used to calculate the influence of a node in the network, it usually requires multiple iterations to converge to a steady-state value, which reflects the long-term influence of the node; the source node set refers to the set of all nodes connected to the current node. For a given node, its source node set contains all other nodes connected to it by edges.

[0076] In an optional embodiment, the heat prediction model employs an LSTM network, and the formula for calculating the spread probability of hot events is as follows:

[0077]

[0078] Among them, P spread (c i ) represents the diffusion probability of a hot topic event, β represents the preset density coefficient, γ represents the preset decay factor, and S iThis indicates the overall social relevance of trending events. This indicates the future popularity characteristics.

[0079] In an optional embodiment, the formula for calculating the caching utility value of hot events is as follows:

[0080]

[0081] Among them, U i This represents the caching utility value of hot events, α represents the preset weighting factor, and Size(c i ) indicates the data size of hot events.

[0082] The embodiments of the present invention have been described in detail above with reference to the accompanying drawings. However, the present invention is not limited thereto. Various changes can be made within the scope of knowledge possessed by those skilled in the art without departing from the spirit of the present invention.

Claims

1. A social caching retrieval system integrating hotspot analysis, characterized in that, include: A heat modeling unit (1) is used to obtain a set of hot events and perform heat modeling on each hot event in the set of hot events to obtain the heat characteristics of the hot events. Social association unit (2), the social association unit (2) is used to obtain a user set, construct a user-content bipartite graph based on the interaction behavior between the hot event set and the user set, and determine the global social association degree of each hot event in the hot event set based on the user-content bipartite graph; The propagation calculation unit (3) is used to predict the popularity characteristics of the hot events based on the pre-trained popularity prediction model, so as to obtain the future popularity characteristics of the hot events, and to propagate the content of the future popularity characteristics of the hot events according to the infectious disease model, so as to obtain the diffusion probability of the hot events; the caching calculation unit (4) is used to define a caching utility function based on the global social correlation degree, future popularity characteristics and diffusion probability of the hot events, and to determine the caching utility value of each hot event in the hot event set according to the caching utility function; The system further includes: a social caching unit (5), which is used to determine a list of hot events and a list of non-hot events based on the cache utility value of the hot events, store the list of hot events in a local cache, and store the list of non-hot events in each cache node in the cache cluster; a retrieval request unit (6), which is used to obtain a retrieval request, match the retrieval request with the hot events in the local cache, and if no match is found, send a query request to the cache node with the highest cache utility value within a preset number of hops according to a distributed hash table; the calculation formula for the user's social influence weight is as follows: ;in, Represents a set of users The Middle individual users The first in the set of hot events Hot topics Social influence weight Represents the first in the user set individual users The first in the set of hot events Hot topics The number of reposts, Represents a set of users The first in the set of hot events Hot topics The total number of reposts; the formula for calculating the popularity characteristics of the hot topic is as follows: ;in, Indicates a time sliding window The characteristics of the popularity of domestic hot topics This indicates adjustable optimization weights. This indicates the user's response to the first event in the set of trending events. Hot topics Number of visits to trending events This represents the total number of times a user accesses the hot events in the hot event set. Indicates the user is in a time-sliding window The first in the internal hot topic event set Hot topics The intensity of interactive behavior, This represents the function for finding the maximum value; determining the global social relevance of each hot event in the hot event set based on the user-content bipartite graph includes: normalizing the final steady-state influence of each node based on the maximum and minimum values ​​of the final steady-state influence of all nodes in the user-content bipartite graph to obtain the global social relevance of the node, i.e., obtaining the global social relevance of the hot event; the popularity prediction model uses an LSTM network, and the formula for calculating the diffusion probability of the hot event is as follows: ;in, Indicates the probability of a hot topic spreading. This represents the preset density coefficient. This represents the preset attenuation factor. This indicates the overall social relevance of trending events. The formula for calculating the caching utility value of the hot events is as follows: This represents the future popularity characteristics. ;in, This represents the caching utility value of hot events. This represents the preset weighting factor. This indicates the size of the data for trending events.

2. The social caching retrieval system integrating hotspot analysis according to claim 1, characterized in that, To obtain the popularity characteristics of each hot event in the hot event set, the following steps are performed: 1) Count the number of visits to the hot event according to a preset time sliding window; 2) Obtain the total number of reposts of the hot events in the user set according to the preset time sliding window, and determine the social influence weight of the user based on the total number of reposts of the hot events in the user set; 3) Obtain the user's interaction behavior within the time sliding window, and define the intensity of the user's interaction behavior within the time sliding window, wherein the interaction behavior includes likes, comments, and reposts, and the intensity of the interaction behavior is a weighted sum of the interaction behaviors; 4) Determine the popularity characteristics of the hot event based on the number of visits to the hot event, the social influence weight of the user interacting with the hot event, and the intensity of the user's interaction behavior.

3. The social caching retrieval system integrating hotspot analysis according to claim 2, characterized in that, The user-content bipartite graph includes a set of user nodes, a set of hot event nodes, a set of connecting edges, and a set of edge weights. The user nodes correspond to users, and the hot event nodes correspond to hot events. The connecting edges are used to connect the user nodes and the hot event nodes. When a user interacts with a hot event, there is a connecting edge between the corresponding user node and the hot event node. The edge weights of the connecting edges are defined according to the intensity of the interaction between the user and the hot event.

4. The social caching retrieval system integrating hotspot analysis according to claim 3, characterized in that, Determining the global social relevance of each hot event in the hot event set based on the user-content bipartite graph further includes: assigning an initial influence value to each node in the user-content bipartite graph, wherein the calculation formula for the initial influence value is as follows: ;in, This represents the first element in the user-content bipartite graph. The initial influence value of each node. This represents the number of nodes in the user-content bipartite graph. The steady-state influence of each node in the user-content bipartite graph is calculated using random walk simulation, where the formula for calculating the steady-state influence is as follows: ;in, Indicates the first The steady-state influence of nodes in the next iteration. Let represent the damping factor, and let the damping factor represent the probability that the user continues the random walk. The remaining... Let be the probability of randomly jumping to any node. This represents the set of source nodes of a node. Represents a node arrive The transition probability, and ， Represents nodes in a user-content bipartite graph. To the node The edge weights are calculated; the above operation is repeated until the steady-state influence after two iterations is less than the preset influence threshold, so as to obtain the final influence threshold.

Images