A social network user emotion propagation model based on game theory

Abstract

The application discloses a social network user emotion propagation model based on game theory, and aims at solving the technical defects that emotion propagation models cannot accurately depict real emotion propagation processes in existing emotion analysis and propagation simulation work. The technical scheme comprises the following steps: performing emotion analysis on a target user node according to a preset emotion analysis method, and extracting emotion labels of the node; constructing an emotion game payoff matrix according to the emotion labels of each node; adjusting a basic propagation probability in an SIR infectious disease model according to the emotion game payoff matrix; adjusting a basic immunity probability in the SIR infectious disease model according to a social reinforcement effect; adjusting a structure of the SIR infectious disease model according to a forgetting mechanism; and obtaining the social network user emotion propagation model. The application is suitable for application in emotion analysis and propagation simulation work.

Description

Technical Field

[0001] This involves the field of social network analysis technology, specifically a social network user sentiment propagation model based on game theory. Background Technology

[0002] In the field of social network analysis, the study of sentiment propagation has received increasing attention in recent years. With the rapid development of the internet and social media, various online platforms have gradually replaced traditional media, becoming the main channels for people to express emotions and share information. Through these platforms, users can communicate and express their emotions and opinions in real time, thus forming complex dynamics of sentiment propagation on social networks. Existing research mainly focuses on the following aspects:

[0003] First, text-based sentiment analysis methods have been widely applied. Researchers use natural language processing techniques, along with dictionaries and machine learning models, to extract the sentiment tendencies in user-posted content. For example, the sentiment dictionary approach is widely used, constructing dictionaries that include words representing positive and negative sentiments to classify and score text on social networks, thereby analyzing the dynamic characteristics of sentiment propagation. Furthermore, machine learning models, especially deep learning techniques, are also used to train sentiment classifiers to achieve higher accuracy in sentiment recognition.

[0004] Secondly, the construction of emotion transmission models is also an important direction in existing research. Commonly used models include the SIR (Susceptibility-Infection-Recovery) model and its variants. These models were originally used to describe the transmission process of infectious diseases, but they have also been introduced into emotion transmission research to simulate the spread of emotions in social networks. However, traditional SIR models mainly focus on calculating the probability of transmission and do not fully consider the emotional tendencies between individuals and the complexity of their transmission decisions. This leads to a certain deviation between the simulation results of emotion transmission and the actual situation.

[0005] While existing research has made significant efforts in sentiment analysis and communication simulation, it often overlooks the decision-making factors individuals consider when disseminating information during the emotional transmission process. For example, when deciding whether to spread a particular message, individuals are influenced not only by the message's inherent emotional tone but also by the emotional similarity between themselves and other users, as well as the potential benefits or costs of dissemination. This decision-making process has a game-like nature, and traditional models lack the ability to characterize this game-like behavior, thus limiting their applicability in complex social environments.

[0006] Therefore, the problem with existing technologies is that they lack consideration of the emotional game between users, which makes the emotional transmission model unable to accurately depict the real emotional transmission process, especially when individual transmission decisions depend on the game of interests. Summary of the Invention

[0007] To address the shortcomings of existing technologies, such as the lack of consideration for emotional game theory among users in current sentiment analysis and propagation simulation, which leads to the inaccurate characterization of real-world sentiment propagation models, this invention provides the following technical solution:

[0008] A game theory-based model for the propagation of user sentiment on social networks, the model construction method includes:

[0009] The steps are: to perform sentiment analysis on target user nodes according to a preset sentiment analysis method and extract the sentiment tags of the nodes;

[0010] The steps for constructing an emotional game payoff matrix based on the emotional tags of each node;

[0011] The steps for adjusting the base transmission probability in the SIR infectious disease model based on the emotional game payoff matrix;

[0012] The step of adjusting the baseline immunity probability in the SIR infectious disease model based on the social reinforcement effect;

[0013] The steps to adjust the structure of the SIR infectious disease model based on the forgetting mechanism to obtain the social network user sentiment propagation model are as follows.

[0014] Furthermore, a preferred implementation is provided, in which an emotional game payoff matrix is ​​constructed based on the emotional tags of each node and the relationship between the emotional tendencies of the information disseminator and the information receiver.

[0015] Furthermore, a preferred embodiment is provided, wherein the preset sentiment analysis method specifically comprises:

[0016] Create an emotion dictionary and assign different emotion scores or labels with different emotion tendencies to the emotion words in the dictionary.

[0017] Furthermore, a preferred embodiment is provided in which the adjustment of the immune probability in the social reinforcement effect is based on the proportion of the immune status of surrounding nodes, and the adjustment magnitude is adjusted by controlling parameters.

[0018] Furthermore, a preferred implementation is provided, wherein the propagation probability adjustment in the SIR model is based on the calculation result of the game equilibrium point in the game payoff matrix between the propagating node and the receiving node.

[0019] Furthermore, a preferred implementation is provided in which the calculation of the game equilibrium point is based on the Bayesian Nash equilibrium condition.

[0020] Based on the same inventive concept, this invention also provides a device for constructing a social network user sentiment propagation model based on game theory, comprising:

[0021] This module performs sentiment analysis on target user nodes based on a preset sentiment analysis method and extracts the sentiment tags of the nodes.

[0022] A module for constructing an emotional game payoff matrix based on the emotional tags of each node;

[0023] A module that adjusts the base transmission probability in the SIR infectious disease model based on the emotional game payoff matrix;

[0024] A module that adjusts the baseline immunity probability in the SIR infectious disease model based on the social reinforcement effect;

[0025] The structure of the SIR infectious disease model is adjusted based on the forgetting mechanism to obtain a module for the social network user sentiment propagation model.

[0026] Based on the same inventive concept, the present invention also provides a computer storage medium for storing a computing program, wherein when the computer reads the computer program, the computer executes the method described thereon.

[0027] Based on the same inventive concept, the present invention also provides a computer, including a processor and a storage medium, wherein when the processor reads a computer program stored in the storage medium, the computer executes the method described thereon.

[0028] Based on the same inventive concept, the present invention also provides a computer program product, which, when executed, implements the method described.

[0029] Compared with the prior art, the advantages of the technical solution provided by the present invention are as follows:

[0030] Introducing an emotional game payoff matrix allows the model to more accurately depict the tradeoffs between the benefits and costs of information dissemination for users during the communication process. This payoff matrix considers the emotional alignment and inconsistency between the disseminating and receiving nodes, and quantifies the benefits of dissemination in different scenarios. Compared to traditional models, it can more realistically simulate users' communication decisions in complex social networks.

[0031] The introduction of social reinforcement improves the model's simulation performance in complex networks. By dynamically adjusting the number of immune nodes surrounding a node, the model can better describe the mutual influence between individuals during emotion propagation. This mechanism makes the adjustment of propagation probability more consistent with the real-world patterns of emotion propagation, especially when social connections between individuals have a significant impact on emotion propagation.

[0032] The application of the forgetting mechanism effectively improves the accuracy and stability of the model in depicting emotion propagation. In the traditional SIR model, infected nodes remain active indefinitely. However, with the introduction of the forgetting mechanism, nodes may forget the information they have received and transition to an immune state with a certain probability. This mechanism makes the model more closely resemble the real-world process of user emotional changes, better describing the decay effect of emotion propagation, especially as negative emotions gradually subside.

[0033] Compared to existing models that rely solely on propagation probability and infection state transitions, this approach significantly improves the accuracy and applicability of the model by introducing emotional game theory, social reinforcement effects, and forgetting mechanisms. These mechanisms work together to enable the model not only to more accurately predict the propagation path of emotions in networks but also to better reflect individual emotional propagation behaviors within social networks.

[0034] Suitable for use in sentiment analysis and communication simulation. Attached Figure Description

[0035] Figure 1 A flowchart illustrating a method for constructing a social network user sentiment propagation model based on game theory;

[0036] Figure 2 This is a schematic diagram illustrating the adjustment of the SIR infectious disease model structure. Detailed Implementation

[0037] To make the advantages and benefits of the technical solution provided by the present invention clearer, the technical solution provided by the present invention will now be described in further detail with reference to the accompanying drawings. Specifically:

[0038] Implementation Method 1: This implementation method provides a social network user sentiment propagation model based on game theory. The model construction method includes:

[0039] The steps are: to perform sentiment analysis on target user nodes according to a preset sentiment analysis method and extract the sentiment tags of the nodes;

[0040] The steps for constructing an emotional game payoff matrix based on the emotional tags of each node;

[0041] The steps for adjusting the base transmission probability in the SIR infectious disease model based on the emotional game payoff matrix;

[0042] The step of adjusting the baseline immunity probability in the SIR infectious disease model based on the social reinforcement effect;

[0043] The steps to adjust the structure of the SIR infectious disease model based on the forgetting mechanism to obtain the social network user sentiment propagation model are as follows.

[0044] Specifically, including:

[0045] Step 1: Sentiment Analysis and Tag Extraction

[0046] First, an improved dictionary-based sentiment analysis method is used to perform sentiment analysis on user nodes in social networks. Specifically, a sentiment dictionary is constructed, including commonly used word segmentation dictionaries, stop word dictionaries, and sentiment word dictionaries. To improve the accuracy of sentiment analysis, seven degree word dictionaries, one negation word dictionary, and two sentiment word dictionaries (positive and negative sentiment word dictionaries, respectively) are also introduced. During the analysis, the text content of user nodes is segmented, stop words are removed, and sentiment words in the text are identified. Sentiment words are matched based on the sentiment dictionary, and sentiment scores are calculated. The sentiment tendency of a node is determined based on the highest-scoring or most frequent tag, and the output is the node's sentiment tag.

[0047] In this process, each text is analyzed based on different sentiment words from the dictionary, with words of varying degrees weighting the sentiment intensity of each word. For example, "very much like" has a stronger sentiment intensity than "like," and therefore receives a higher weight in the analysis. Negative words also correct the sentiment tendency of sentiment words; for instance, "dislike" reverses the sentiment of "like" to a negative one. Through these steps, the sentiment analysis results can more accurately reflect the user's true emotional state.

[0048] Step 2: Constructing the payoff matrix for emotional game theory

[0049] Based on the user node sentiment tags extracted in Step 1, a sentiment game payoff matrix is ​​constructed. This payoff matrix considers different scenarios where the sentiments of the propagating node and the receiving node align, and quantifies the propagation payoff in each situation. For example, when the sentiments of the propagating node and the receiving node align and both choose to propagate, both parties receive high payoffs; however, when their sentiments clash, the propagating node may face penalties. The output of this step is the sentiment game payoff matrix, which is used to model the propagation process in subsequent steps.

[0050] When constructing the payoff matrix, node choices under various scenarios must be considered. For example, nodes can choose to propagate or not propagate information, and receiving nodes can choose to accept or not accept information. When sentiments align, both propagation and reception bring positive payoffs, while when sentiments clash, propagation may result in penalties or negative payoffs. This payoff setting can simulate the decision-making behavior of individuals in the process of sentiment propagation in social networks, thereby more accurately predicting the propagation outcome.

[0051] Step 3: Integrating the emotional game payoff matrix with the SIR model

[0052] The sentiment game payoff matrix constructed in step two is then integrated with the SIR infectious disease model. The basic propagation probability in the SIR model is dynamically adjusted using the sentiment game payoff matrix, ensuring that the propagation probability is influenced not only by node states but also by the sentiment game relationship between the propagating and receiving nodes. Specifically, the propagation probability is optimized using Bayesian Nash equilibrium conditions to better simulate the propagation process of sentiment in the network.

[0053] In this process, the emotional states and payoff matrices of the propagating and receiving nodes determine the probability of information propagation. If the payoff from propagation is high, the probability of propagation increases accordingly, and vice versa. By introducing game theory, the model can reflect the rational choices individuals make during the propagation process, simulating propagation dynamics that more closely resemble real-world scenarios.

[0054] Step 4: Introduce social reinforcement effect to adjust the probability of immunity

[0055] Building upon the SIR model, a social reinforcement effect is introduced to adjust the baseline immunity probability. The immunity probability of a node is dynamically adjusted by considering the number of immune nodes surrounding it. Specifically, as the number of surrounding immune nodes increases, the node's immunity probability also increases. When there are no surrounding immune nodes, the immunity probability equals the baseline immunity probability. This dynamic adjustment method makes the model more closely resemble the immunity behavior of individuals in real-world social networks.

[0056] Social reinforcement can describe an individual's conformity behavior in social networks. When a node is surrounded by a large number of immune nodes, the likelihood of that node choosing not to spread the virus or remaining immune increases. This mechanism can better describe the phenomenon of emotional transmission inhibition caused by social pressure or conformity in reality, thereby improving the realism and accuracy of the model.

[0057] Step 5: Introduce a forgetting mechanism to adjust the SIR model

[0058] Introducing a forgetting mechanism into the traditional SIR model allows infected nodes to forget the received information and transition to an immune state with a certain probability. As the forgetting probability increases, the number of infected nodes becoming immune nodes increases, effectively describing the decay effect of emotion propagation. Introducing the forgetting mechanism allows the model to better simulate the real-world emotion fading process after emotion propagation ends, making it particularly suitable for scenarios where negative emotions gradually fade.

[0059] The introduction of the forgetting mechanism allows the model to dynamically simulate the termination process of emotion transmission. In real-world social networks, negative emotions tend to fade gradually over time, a phenomenon reflected through the forgetting mechanism. As the forgetting probability changes, the model can describe the persistence and rate of fading of emotions under different circumstances, thus providing a theoretical basis for managing and intervening in the transmission of negative emotions.

[0060] Implementation Method Two: This implementation method further defines the game theory-based social network user sentiment propagation model provided in Implementation Method One. Based on the sentiment tags of each node and the relationship between the sentiment inclinations of the information disseminator and the sentiment inclinations of the information receiver, an sentiment game payoff matrix is ​​constructed.

[0061] Implementation Method 3: This implementation method further defines the game theory-based social network user sentiment propagation model provided in Implementation Method 1. The preset sentiment analysis method is specifically as follows:

[0062] Create an emotion dictionary and assign different emotion scores or labels with different emotion tendencies to the emotion words in the dictionary.

[0063] Implementation Method Four: This implementation method is a further limitation of the game theory-based social network user sentiment propagation model provided in Implementation Method One. In the social reinforcement effect, the adjustment of the immunity probability is based on the proportion of the immunity status of surrounding nodes, and the adjustment range is adjusted by controlling parameters.

[0064] Implementation Method 5: This implementation method is a further limitation of the social network user sentiment propagation model based on game theory provided in Implementation Method 1. The propagation probability adjustment in the SIR model is based on the calculation result of the game equilibrium point in the game payoff matrix between the propagating node and the receiving node.

[0065] Implementation Method Six: This implementation method is a further limitation of the game theory-based social network user sentiment propagation model provided in Implementation Method One. The calculation of the game equilibrium point is based on the Bayesian Nash equilibrium condition.

[0066] Implementation Method Seven: This implementation method provides a device for constructing a social network user sentiment propagation model based on game theory, comprising:

[0067] This module performs sentiment analysis on target user nodes based on a preset sentiment analysis method and extracts the sentiment tags of the nodes.

[0068] A module for constructing an emotional game payoff matrix based on the emotional tags of each node;

[0069] A module that adjusts the base transmission probability in the SIR infectious disease model based on the emotional game payoff matrix;

[0070] A module that adjusts the baseline immunity probability in the SIR infectious disease model based on the social reinforcement effect;

[0071] The structure of the SIR infectious disease model is adjusted based on the forgetting mechanism to obtain a module for the social network user sentiment propagation model.

[0072] Implementation Method 8: This implementation method provides a computer storage medium for storing a computing program. When the computer reads the computer program, the computer executes the method provided in Implementation Method 1.

[0073] Implementation Method Nine: This implementation method provides a computer, including a processor and a storage medium. When the processor reads a computer program stored in the storage medium, the computer executes the method provided in Implementation Method One.

[0074] Implementation Method 10: This implementation method provides a computer program product. As a computer program, when the computer program is executed, it implements the method provided in Implementation Method 1.

[0075] Implementation Method Eleven: Combination Figure 1 and 2 This embodiment describes the technical solution provided above in further detail through specific examples. Specifically:

[0076] Reference Figure 1 As shown, this embodiment proposes an example of a social network user sentiment propagation model based on game theory. The construction method includes the following steps:

[0077] Step 1: Perform sentiment analysis on user nodes using an improved dictionary-based sentiment analysis method to extract the sentiment tags of the nodes;

[0078] Step 2: Based on the sentiment tags of each node, construct the sentiment game payoff matrix according to the relationship between the sentiment inclination of the information disseminator and the sentiment inclination of the information receiver;

[0079] Step 3: Integrate the emotional payoff game matrix with the SIR infectious disease model, and adjust the basic transmission probability in the infectious disease model through the emotional payoff matrix;

[0080] Step 4: Utilize the social reinforcement effect to adjust the baseline immunization probability in the transmission model;

[0081] Step 5: Adjust the structure of the SIR infectious disease model using the forgetting mechanism to complete the construction of a social network user sentiment propagation model based on game theory.

[0082] In step one, an improved dictionary-based sentiment analysis method is used to perform sentiment analysis on user nodes. Specifically, the dictionary-based sentiment analysis method first establishes a sentiment dictionary. In constructing the sentiment dictionary, in addition to the commonly used word segmentation dictionary, stop word dictionary, and sentiment word dictionary, seven degree word dictionaries, one negation word dictionary, and two sentiment word dictionaries (one positive sentiment word dictionary and one negative sentiment word dictionary) are added to increase the accuracy of sentiment analysis. Different sentiment scores or different sentiment tendency labels are assigned to the sentiment words in the dictionary. Then, in the sentiment analysis process, the text of the user node is segmented. Stop words in the text are removed based on the word segmentation dictionary and the stop word dictionary to find the sentiment words in the text. The sentiment words in the text are directly matched with the sentiment words in the sentiment dictionary. Based on the sentiment scores or sentiment tendency labels of all words in the text, the label with the highest score or the largest number is selected as the final sentiment tendency of the text.

[0083] Specifically, when performing sentiment analysis on the sentence, "I especially like Wuhan! Because Wuhan has many beautiful scenic spots. However, I don't like the weather in Wuhan because it's a bit bad; the heat makes me feel uncomfortable," to analyze the sentiment tendency of a sentence, we first need to identify the sentiment words, degree adverbs, and negative words in the sentence. Positive sentiment words include "beautiful" and "like," while negative sentiment words include "bad" and "unpleasant." Each occurrence of a positive word adds 1 point to the score, and each occurrence of a negative word subtracts 1 point. The sentiment words "like," "beautiful," and "bad" are all preceded by a degree adverb. "Excellent" is more emotionally intense than "good" or "relatively good," and "terrible" is stronger than "somewhat bad." Therefore, after identifying the sentiment words, we need to look ahead for degree adverbs (different degree adverbs have different weights here). We categorized degree words into six types: most, very, more, ish, insufficient, and over. Through testing and calculation, the scores for these six sentiment words were 4, 3, 2, 0.5, -0.3, and -0.5, respectively. The sentence "I especially like the city of Wuhan" is followed by an exclamation mark, indicating strong emotion. Therefore, the exclamation mark adds 2 to the sentiment score. Furthermore, when finding sentiment words, we need to look for negative words preceding them, such as "no" and "cannot." We also need to count the frequency of these negative words; if the frequency is odd, the sentiment score is multiplied by 1, but if it's even, the sentiment remains unchanged. In this sentence, we can see that there's only one "no" before "like," so the sentiment value for "like" should be reversed, multiplied by 1. In this implementation method, positive and negative sentiments are calculated separately because a text contains both praise and criticism, and a single score cannot represent its sentiment tendency. By scoring each clause and then weighted and summing the scores, we obtain a final positive score and a negative score. The higher score indicates the sentiment tendency of the text. In this example, the final score of the five clauses is [positive score, negative score] = [8, 1.5]. Since 8 > 1.5, the sentiment of the whole sentence is positive. Therefore, sentiment analysis can be performed on the text of the node to extract the sentiment label.

[0084] In step two, the payoff matrix of the emotional game is constructed. The specific construction method is as follows:

[0085] Game payoff matrix

[0086]

[0087] In this context, the cost of a propagating node propagating information to another node is *c*; the benefit for both nodes when successfully propagating a sentimentally compatible message is *x*; the penalty for a propagating node successfully propagating a sentimentally incompatible message is *c*; and the benefit for both nodes when preventing the propagation of a sentimentally incompatible message is *d*. When the sentiments of the propagating and receiving nodes are compatible, if the propagating node chooses to propagate and the receiving node chooses to receive, the receiving node's benefit is *x*, and the propagating node's benefit is *xc*. When the sentiments of the propagating and receiving nodes are compatible, if the propagating node chooses to propagate and the receiving node chooses not to receive, the receiving node's benefit is 0, and the propagating node's benefit is -*c*. When the sentiments of the propagating and receiving nodes are compatible, if the propagating node chooses not to propagate, both the receiving and propagating nodes' benefits are 0. When the sentiments of the propagating and receiving nodes are incompatible, if the propagating node chooses to propagate and the receiving node chooses to receive... The payoff for the receiving node is -c, and the payoff for the propagating node is -2c. When the sentiments of the propagating and receiving nodes are inconsistent, if the propagating node chooses to propagate and the receiving node chooses not to receive, the payoff for the receiving node is d, and the payoff for the propagating node is -2c. When the sentiments of the propagating and receiving nodes are inconsistent, if the propagating node chooses not to propagate and the receiving node chooses to receive, the payoff for the receiving node is 0, and the payoff for the propagating node is dc. When the sentiments of the propagating and receiving nodes are inconsistent, if the propagating node chooses not to propagate and the receiving node chooses not to receive, the payoff for the receiving node is d, and the payoff for the propagating node is dc. In step three, the basic transmission probability in the infectious disease model is adjusted using the sentiment game payoff matrix. The specific adjustment method is as follows:

[0088] p·x=(1-p)·d

[0089] The Bayesian Nash equilibrium of the game payoff matrix exists under the conditions that x>c>0 and dc>0, and its Bayesian Nash equilibrium is shown in the above formula, which can be transformed into the following formula:

[0090]

[0091] Where p is the adjusted propagation probability, x is the reward for both nodes when a sentiment-matching message is successfully propagated, and d is the reward for preventing the propagation of a sentiment-disagreeing message.

[0092] In step four, the social reinforcement effect is used to adjust the baseline immunity probability in the transmission model. The specific adjustment method is as follows:

[0093] γ=γ base ·(log a( b·n)+1)

[0094] Where γ is the adjusted immune probability, γbase is the base immune probability of the propagation model, n is the number of immune nodes around the node, and a and b are control parameters. The degree of immune probability adjustment can be controlled by controlling the values ​​of parameters a and b. As the number of immune nodes n around the node increases, the immune probability γ will increase on the basis of the base immune probability γbase. When n = 0, the immune probability is equal to the base immune probability.

[0095] In step five, the structure of the SIR infectious disease model is adjusted using a forgetting mechanism, specifically as follows: Figure 2 As shown.

[0096] A forgetting mechanism is introduced into the traditional SIR model. Infected nodes forget the information they have received and thus become immune. As the forgetting probability increases, the number of infected nodes that become immune nodes increases, resulting in a larger number of immune nodes after propagation compared to the model without the forgetting mechanism. Conversely, as the forgetting probability decreases, the number of infected nodes that become immune nodes still increases, resulting in a larger number of immune nodes after propagation compared to the model without the forgetting mechanism, but the increase is less than when the forgetting probability is higher.

[0097]

[0098] The above formula is the propagation dynamics formula for a social network user sentiment propagation model based on game theory, where β is the propagation probability, γ is the immunity probability, and η is the forgetting probability. Therefore, the larger β is, the greater the probability of transmission from state S to state I, and the more S nodes become state I. Similarly, the larger γ is, the greater the probability of immunity from state I to state R, and the more I nodes become state R.

[0099] The above description of several specific embodiments further details the technical solution provided by the present invention in order to highlight the advantages and benefits of the technical solution provided by the present invention. However, the above-described specific embodiments are not intended to limit the present invention. Any reasonable modifications and improvements to the present invention, combinations of embodiments, and equivalent substitutions based on the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A social network user sentiment propagation model based on game theory, characterized in that, Model building methods include: The steps are: to perform sentiment analysis on target user nodes according to a preset sentiment analysis method and extract the sentiment tags of the nodes; The steps for constructing an emotional game payoff matrix based on the emotional tags of each node; The steps for adjusting the base transmission probability in the SIR infectious disease model based on the emotional game payoff matrix; The step of adjusting the baseline immunity probability in the SIR infectious disease model based on the social reinforcement effect; The steps for adjusting the structure of the SIR infectious disease model based on the forgetting mechanism to obtain the social network user sentiment propagation model are as follows: The emotional game payoff matrix considers the different situations of emotional consistency and inconsistency between the propagation node and the receiving node, and quantifies the basic propagation payoff in each scenario; the output of this step is the emotional game payoff matrix, which is used to model the propagation process in subsequent steps. By integrating the emotional game payoff matrix with the SIR infectious disease model, the basic transmission probability in the SIR model is dynamically adjusted through the emotional game payoff matrix, so that the transmission probability is not only affected by the node state, but also depends on the emotional game relationship between the propagating node and the receiving node. The adjustment of the basic propagation probability in the SIR model is based on the calculation result of the game equilibrium point in the game payoff matrix between the propagation node and the receiving node; the calculation of the game equilibrium point is based on the Bayesian Nash equilibrium condition. Based on the SIR model, the social reinforcement effect is introduced to adjust the baseline immunization probability; pass: The above formula is the propagation dynamics formula of the social network user sentiment propagation model based on game theory, where β is the propagation probability, γ is the immunity probability, and η is the forgetting probability. When β is larger, the probability of being infected from state S to state I is greater, and the number of S nodes that become state I is greater. When γ is larger, the probability of being immune from state I to state R is greater, and the number of I nodes that become state R is greater. Introducing a forgetting mechanism into the traditional SIR model allows infected nodes to forget the received information and transition to an immune state with a certain probability. An improved dictionary-based sentiment analysis method is used to perform sentiment analysis on user nodes in social networks. Specifically, a sentiment dictionary is constructed, which includes commonly used word segmentation dictionaries, stop word dictionaries, and sentiment word dictionaries. Seven degree word dictionaries, one negation word dictionary, and positive and negative sentiment word dictionaries are introduced. During the analysis, the text content of user nodes is segmented, stop words are removed, and sentiment words in the text are identified. Sentiment words are matched based on the sentiment dictionary, and sentiment scores are calculated. The sentiment tendency of the node is determined based on the tag with the highest score or the most numerous tags, and the output is the sentiment tag of the node.

2. A device for constructing a social network user sentiment propagation model based on game theory, characterized in that, To implement the method of claim 1, the method comprises: This module performs sentiment analysis on target user nodes based on a preset sentiment analysis method and extracts the sentiment tags of the nodes. A module for constructing an emotional game payoff matrix based on the emotional tags of each node; A module that adjusts the base transmission probability in the SIR infectious disease model based on the emotional game payoff matrix; A module that adjusts the baseline immunity probability in the SIR infectious disease model based on the social reinforcement effect; The structure of the SIR infectious disease model is adjusted based on the forgetting mechanism to obtain a module for the social network user sentiment propagation model.

3. A computer storage medium for storing computing programs, characterized in that, When the computer reads the computer program, the computer executes the method of claim 1.

4. A computer, including a processor and a storage medium, characterized in that, When the processor reads the computer program stored in the storage medium, the computer executes the method of claim 1.

5. A computer program product, as a computer program, is characterized by: When the computer program is executed, it implements the method of claim 1.

Images