Content influence measuring method based on social media

Abstract

The invention relates to the field of influence measurement, and discloses a social media-based content influence measurement method, which comprises the following steps of: constructing a social network graph according to collected user social relation data; each user is a node, interaction between the users is an edge, and features in the social network are analyzed; modeling a propagation path of the information by adopting a propagation model, and weighting the weight of the information propagation according to the social relationship between the users; establishing a weight model, endowing each edge in the propagation path with different weights according to the strength of different social relationships, and strengthening the influence of the core social group on content propagation; evaluating the propagation breadth and propagation speed of the content in the social network, analyzing the participation behavior of the user, and evaluating the propagation effects of the content in different social circles; and training the data by using a machine learning algorithm, and optimizing the propagation effect prediction model. The method has the advantage of accurately evaluating the actual propagation effect and influence of the content.

Description

Technical Field

[0001] The present invention relates to the field of influence measurement, and in particular to a method for measuring content influence based on social media. Background Art

[0002] With the rapid development of social media, it has become a primary platform for daily communication, information dissemination, and opinion expression, generating a vast amount of user-generated content. The dissemination and influence of this content has become a hot topic of research, particularly in areas such as brand marketing, public opinion analysis, and social event monitoring. Understanding the influence of social media content is crucial. The influence of social media content generally refers to the ability of a piece of content or information to spread and trigger reactions online, encompassing aspects such as exposure, user interaction, and reach.

[0003] Traditional influence measurement methods rely primarily on static metrics such as clicks, likes, reposts, and comments. However, these methods fail to fully reflect the actual impact of information because they overlook the network structure, users' social relationships, and their influence during content dissemination. In social media, information dissemination depends not only on the quality of the content itself but also on the structure of the publisher and the social network. Against this backdrop, developing a more accurate content influence measurement method that comprehensively considers the topology of social networks, information dissemination pathways, and user behavior has become an important research direction.

[0004] In recent years, with the development of social network analysis technology, the measurement of the influence of social media content has gradually evolved from a single quantitative indicator to a multi-dimensional assessment. For example, measurement methods based on social network analysis can analyze the dissemination potential of content by analyzing the connections and propagation paths between nodes, taking into account the influence propagation mechanism in information flow. Furthermore, measurement methods based on sentiment analysis and user interaction analysis can more comprehensively assess the emotional influence of content and its impact on users through the analysis of user comments and feedback. Furthermore, the introduction of deep learning and machine learning methods has enabled influence measurement methods to more intelligently and automatically identify high-impact content, improving prediction accuracy and application effectiveness.

[0005] Although many influence measurement methods have been proposed, most still have certain limitations, such as failing to fully consider the impact of social relationships between users on information dissemination. Therefore, it is necessary to design a content influence measurement method based on social media that can accurately evaluate the actual dissemination effect and influence of content. Summary of the Invention

[0006] (1) Technical problems solved In response to the shortcomings of the existing technology, the present invention provides a content influence measurement method based on social media, which has the advantage of accurately evaluating the actual dissemination effect and influence of content, and solves the problems in the above-mentioned background technology.

[0007] (2) Technical solution To achieve the above-mentioned purpose of accurately evaluating the actual dissemination effect and influence of content, the present invention provides the following technical solution: a method for measuring the influence of content based on social media, comprising the following steps: S1: Build a social network graph based on the collected user social relationship data, with each user as a node and the interactions between users as edges, and analyze the characteristics of the social network; Preferably, S1 further includes node feature analysis, which measures the average shortest path distance between a node and all other nodes. Users with high proximity centrality are usually more likely to quickly obtain and spread information. The formula is:

[0008] Where, is the shortest path from node u to node v; It measures the degree to which a node acts as an intermediary between other nodes. Users with high betweenness centrality control important paths for information flow. The formula is:

[0009] Where, is the number of shortest paths from node s to node t, is the number of shortest paths passing through node u; Each edge represents an interaction between users. The intensity of the interaction can be weighted according to the frequency and intimacy of the interaction between users. By counting the number of interactions between users, the strength of the social relationship between each pair of users can be evaluated.

[0010] S2: Use a propagation model to model the information propagation path and weight the information propagation according to the social relationships between users; Preferably, the S2 further includes a weight of the social relationship that reflects the strength, trust, and interaction frequency of the relationship between users, and is weighted by the following method: Interaction frequency: The weight is set according to the frequency of interaction between users. Users who interact frequently have closer relationships and higher communication weight. Relationship type: assign different weights based on the relationship type between users; Social network structure: Combined with the user's position in the social network, users with high centrality have stronger communication capabilities, so their communication weight can be set high; Information dissemination history: The dissemination weight is set based on the success or failure of the user's past dissemination of content; The propagation effect of each edge depends on the edge weight, that is, the strength of the relationship between users. In the propagation model, the recursive propagation algorithm is used to simulate the process of information propagation in the network. Edges with high weights will accelerate information propagation, while edges with low weights will slow it down. In the independent cascade model, users will try to spread information to their neighbors and decide whether to continue spreading based on the probability of successful propagation. The success probability of propagation is weighted based on the strength of social relationships and is calculated as follows:

[0011] Where, is the probability that information propagates from user u to user v, is the edge weight, is the set of neighbor nodes of user u.

[0012] S3: Establish a weighting model to assign different weights to each edge in the communication path based on the strength of different social relationships, thereby strengthening the influence of core social groups on content dissemination; Preferably, the S3 further includes the strength of social relationships between users, including interaction frequency, relationship type, and historical communication effect. The model identifies the core nodes in the network and increases the weight of the communication path between the core nodes. The communication effect of the core social group is amplified in the entire network. The weight model comprehensively reflects the strength of social relationships and improves the influence of core users through weighted communication paths.

[0013] S4: Evaluate the breadth and speed of content dissemination in social networks, analyze user participation behavior, and evaluate the dissemination effect of content in different social circles; Preferably, the S4 further includes that the spread breadth can be measured by calculating the number of affected nodes, that is, how many different user nodes the content has successfully spread to after starting from the source node. The calculation formula for the spread breadth is as follows:

[0014] Where S is the source node of the information, is the probability of propagation from source node S to node v, threshold is the minimum probability of effective propagation, and propagation speed reflects the time required for information to propagate from the source node to the target node, which is measured by the average time of the propagation path. The propagation speed is calculated as follows:

[0015] Where, is the propagation time required for information to reach node v from the source node S, and V is the set of all affected nodes.

[0016] S5: Use machine learning algorithms to train data and optimize the communication effect prediction model. Use historical data to train the model and predict the effect and influence of content dissemination.

[0017] Preferably, the S5 further includes collecting and organizing historical data, including user behavior data, extracting characteristic variables that affect the dissemination effect, and selecting a machine learning algorithm to train the model. During the training process, the goal of the model is to predict the breadth of content dissemination by learning the dissemination patterns in historical data, and the dissemination speed is predicted through a regression model or a time series model. The input features include the length of the dissemination path, user activity, and dissemination time.

[0018] (3) Beneficial effects Compared with the existing technology, the present invention provides a content influence measurement method based on social media, which has the following beneficial effects: The present invention uses a propagation model to model the propagation path of information, weighting the weight of information propagation based on the social relationships between users. A weighting model is established to assign different weights to each edge in the propagation path based on the strength of different social relationships, thereby strengthening the influence of core social groups on content propagation. The method also evaluates the breadth and speed of content propagation in social networks, analyzes user engagement behavior, and assesses the effectiveness of content propagation in different social circles. A machine learning algorithm is used to train data, optimize the propagation effect prediction model, and train the model using historical data to predict the effectiveness and influence of content propagation. This method has the advantage of accurately evaluating the actual propagation effect and influence of content. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 Schematic diagram of the method of the present invention. DETAILED DESCRIPTION

[0020] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0021] Example 1 The present invention provides a technical solution: a method for measuring content influence based on social media, comprising the following steps: S1: Build a social network graph based on the collected user social relationship data, with each user as a node and the interactions between users as edges, and analyze the characteristics of the social network; Each edge represents an interaction between users. The strength of an interaction can be weighted based on factors such as the frequency of interaction and intimacy between users. For example, likes, reposts, and comments can be assigned different weights to reflect the closeness of the relationship between users. By counting the number of interactions between users, such as the frequency of comments, shares, and likes, the strength of the social relationship between each pair of users can be assessed. Users who interact frequently have closer relationships and are more likely to influence information dissemination. Different types of interactions can be distinguished, each of which may represent different levels of dissemination strength or influence. For example, reposts generally indicate a strong willingness to spread, while comments may indicate high engagement.

[0022] By dividing the community in the network, for example, using methods such as modularity or spectral clustering, different social groups or interest groups can be identified. The communication effect within the community is usually stronger because there are closer connections between community members. Using algorithms such as the Louvain algorithm can effectively detect communities in social networks and measure the "six degrees of separation" characteristics of social networks, that is, whether the average shortest path between any two nodes is short. The small-world effect shows that although there are many users in a social network, information can reach more distant nodes in just a few steps, which is very important for content dissemination.

[0023] Measures the ratio of the actual number of connections in a social network to the maximum number of possible connections. A higher network density generally means that information can spread faster because the users in the network are more closely related. The formula is:

[0024] Where D is the network density, is the number of edges in the network, is the number of nodes in the network.

[0025] S2: Use a propagation model to model the information propagation path and weight the information propagation according to the social relationships between users; Different communication models are suitable for different types of social networks and communication situations. Common communication models include: Independent cascade model: It assumes that information propagation is carried out through accidental attempts between nodes. Each node has a certain probability of transmitting information in each propagation attempt, and the success or failure depends on the strength of social relationships and the user's propagation tendency.

[0026] Linear Threshold Model: Assume that each node has a "threshold" and will only spread information when it receives a sufficient number of messages from neighboring nodes. The weight of each edge determines the degree of influence of neighbor information on the node.

[0027] SIR model: A model suitable for simulating the spread of viruses or diseases, which models information dissemination based on the user's infection status.

[0028] By analyzing the propagation path of information from source nodes to target nodes, the effectiveness of content dissemination within a specific area can be evaluated. For example, when information is transmitted from a central user to other users, the dissemination effect may vary depending on the weight of social relationships. The reach and influence of information can be assessed across the entire network. Dissemination effectiveness can be evaluated by calculating the spread range, speed, and depth. Using the Monte Carlo method, multiple dissemination simulations can be performed to estimate the average dissemination effect and optimize dissemination strategies. Monte Carlo simulations can help understand the possible dissemination paths and effects of information under different social relationship weights. Based on the dissemination effect evaluation, the dissemination path can be further optimized. By selecting the optimal user dissemination sequence or screening the most influential users, the efficiency of information dissemination can be improved.

[0029] S3: Establish a weighting model to assign different weights to each edge in the communication path based on the strength of different social relationships, thereby strengthening the influence of core social groups on content dissemination; To build a weighting model that assigns different weights to each edge in a communication path based on the strength of the social relationship, we first need to quantify the social relationships between users. Specifically, the strength of social relationships can be quantified using multiple dimensions: interaction frequency, such as the frequency of likes, comments, and reposts, is used to assess the closeness of user relationships, with edges associated with frequent interactions being given higher weights. Relationship types, such as friends, colleagues, and followers, are assigned different weights using preset coefficients; a friend relationship might be assigned a weight of 1.5, while a follower relationship might be assigned a weight of 0.5. Historical communication effects, such as the success rate of reposts and the breadth of dissemination, are weighted by accumulating dissemination results, with nodes with historical influence receiving higher weights. Trustworthiness is quantified using interactive feedback within social relationships, such as whether an information has been reposted or cited multiple times, with edges associated with high trustworthiness receiving higher weights. Furthermore, core social groups can be identified using social network analysis methods, such as the Louvain community detection algorithm or screening based on centrality metrics. Nodes within core groups, such as highly centralized users, enhance their influence on information dissemination by increasing the edge weights of the communication paths between these nodes.

[0030] S4: Evaluate the breadth and speed of content dissemination in social networks, analyze user participation behavior, and evaluate the dissemination effect of content in different social circles; To assess the reach and speed of content dissemination on social networks, we first need to quantify the number of nodes the content reaches and the time efficiency of dissemination. Dissemination can be measured by calculating the number of affected nodes, that is, the number of different user nodes to which the content successfully spreads after starting from the source node. Specifically, the formula for calculating dissemination is as follows:

[0031] Where S is the source node of the information, is the probability of propagation from source node S to node v, threshold is the minimum probability of effective propagation, and propagation speed reflects the time required for information to propagate from the source node to the target node, which is measured by the average time of the propagation path. The propagation speed is calculated as follows:

[0032] Where, is the propagation time required for information to travel from source node S to node v, and V is the set of all affected nodes. To analyze user engagement, we can quantify the degree of engagement based on user interactions. The spread of content within social circles can be analyzed by grouping users and evaluating the differences in spread within and outside each social circle. For example, by calculating the spread breadth and speed within each social circle, as well as the spread attenuation between circles, we can assess the cross-circle spread of content.

[0033] S5: Use machine learning algorithms to train data and optimize the communication effect prediction model. Use historical data to train the model and predict the effect and influence of content dissemination.

[0034] To use machine learning algorithms to train data and optimize the communication effect prediction model, we first need to collect and organize historical data. This includes user behavior data such as likes, comments, and shares; social relationship data such as interaction frequency and relationship strength; and content characteristics such as content type, release time, and topic popularity. By analyzing this data, we can extract a series of characteristic variables that influence communication effects. Next, we select an appropriate machine learning algorithm to train the model.

[0035] During training, the model aims to predict the spread of content by learning from propagation patterns in historical data. For example, a regression model can be used to predict the spread of content, with the target variable being the number of affected nodes (Spread(S)). Input features include user interaction frequency, social relationship weight, and content characteristics. Propagation speed can be predicted using a regression model or a time series model, with input features including the length of the propagation path, user activity, and duration of the propagation.

[0036] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.

[0037] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A content influence measurement method based on social media, characterized in that: The following steps are involved: S1: Build a social network graph based on the collected user social relationship data, with each user as a node and the interactions between users as edges, and analyze the characteristics of the social network; S2: Use a propagation model to model the information propagation path and weight the information propagation according to the social relationships between users; S3: Establish a weighting model to assign different weights to each edge in the communication path based on the strength of different social relationships, thereby strengthening the influence of core social groups on content dissemination; S4: Evaluate the breadth and speed of content dissemination in social networks, analyze user participation behavior, and evaluate the dissemination effect of content in different social circles; S5: Use machine learning algorithms to train data and optimize the communication effect prediction model. Use historical data to train the model and predict the effect and influence of content dissemination.

2. The method for measuring content influence based on social media according to claim 1, characterized in that: The S1 further includes node feature analysis, which measures the average shortest path distance between a node and all other nodes. Users with high proximity centrality are usually more likely to quickly obtain and spread information. The formula is: Where, is the shortest path from node u to node v; It measures the degree to which a node acts as an intermediary between other nodes. Users with high betweenness centrality control important paths for information flow. The formula is: Where, is the number of shortest paths from node s to node t, is the number of shortest paths passing through node u; Each edge represents an interaction between users. The intensity of the interaction can be weighted according to the frequency and intimacy of the interaction between users. By counting the number of interactions between users, the strength of the social relationship between each pair of users can be evaluated.

3. The method for measuring content influence based on social media according to claim 1, characterized in that: The S2 further includes the weight of social relationships, which reflects the strength, trust, and interaction frequency of the relationship between users, and is weighted by the following method: Interaction frequency: The weight is set according to the frequency of interaction between users. Users who interact frequently have closer relationships and higher communication weight. Relationship type: assign different weights based on the relationship type between users; Social network structure: Combined with the user's position in the social network, users with high centrality have stronger communication capabilities, so their communication weight can be set high; Information dissemination history: The dissemination weight is set based on the success or failure of the user's past dissemination of content; The propagation effect of each edge depends on the edge weight, that is, the strength of the relationship between users. In the propagation model, the recursive propagation algorithm is used to simulate the process of information propagation in the network. Edges with high weights will accelerate information propagation, while edges with low weights will slow it down. In the independent cascade model, users will try to spread information to their neighbors and decide whether to continue spreading based on the probability of successful propagation. The success probability of propagation is weighted based on the strength of social relationships and is calculated as follows: Where, is the probability that information propagates from user u to user v, is the edge weight, is the set of neighbor nodes of user u.

4. The method for measuring content influence based on social media according to claim 1, characterized in that: The S3 further includes the strength of social relationships between users, including interaction frequency, relationship type, and historical communication effects. The model identifies core nodes in the network and increases the weight of the communication paths between core nodes. The communication effect of the core social group is amplified in the entire network. The weight model comprehensively reflects the strength of social relationships and improves the influence of core users through weighted communication paths.

5. The method for measuring content influence based on social media according to claim 1, characterized in that: S4 further includes that the spread breadth can be measured by calculating the number of affected nodes, that is, how many different user nodes the content has successfully spread to after starting from the source node. The calculation formula for the spread breadth is as follows: Where S is the source node of the information, is the probability of propagation from source node S to node v, threshold is the minimum probability of effective propagation, and propagation speed reflects the time required for information to propagate from the source node to the target node, which is measured by the average time of the propagation path. The propagation speed is calculated as follows: Where, is the propagation time required for information to reach node v from the source node S, and V is the set of all affected nodes.

6. The method for measuring content influence based on social media according to claim 1, characterized in that: The S5 further includes collecting and organizing historical data, including user behavior data, extracting characteristic variables that affect the dissemination effect, and selecting a machine learning algorithm to train the model. During the training process, the goal of the model is to predict the breadth of content dissemination by learning the dissemination patterns in historical data, and the dissemination speed is predicted through a regression model or a time series model. The input features include the length of the dissemination path, user activity and dissemination time.

Images