Awareness-epidemic disease co-evolution transmission analysis method fusing self-media influence

Abstract

The invention discloses a self-media influence-fused consciousness-epidemic disease co-evolution transmission analysis method, which aims at the limitation that the existing epidemic disease model does not fully consider the key influence of self-media on public consciousness awakening and epidemic disease transmission. Through crawling and integrating social media and disease prevention and control data, a double-layer coupling network model of consciousness-epidemic disease co-evolution propagation under a self-media background is constructed; wherein the upper-layer virtual contact network realizes consciousness diffusion through interaction between a common individual and a self-media node, and the lower-layer physical contact network nodes are in one-to-one correspondence with the upper-layer virtual contact network to simulate an infectious disease transmission process; a microcosmic Markov chain method is proposed to analyze a model dynamics process, and the influence of self-media propaganda guidance and individual awareness on epidemic disease outbreak is quantitatively analyzed by establishing a state transition equation set and deducing a critical threshold value under a steady state condition; according to the invention, a quantitative analysis tool is provided for propagation law research and accurate intervention control of sudden epidemic diseases in a we-media environment.

Description

A Consciousness-Epidemic Co-evolutionary Dissemination Analysis Method Integrating the Influence of Social Media Technical Field

[0001] This invention belongs to the field of public health communication analysis technology, specifically involving a co-evolutionary communication analysis method for consciousness-epidemic influence that integrates the impact of social media. Background Technology

[0002] From the past devastating malaria, dengue fever, and plague to recent outbreaks of avian influenza and Severe Acute Respiratory Syndrome (SARS), the rapid development of modern social media has effectively heightened public awareness of disease prevention and led to a series of self-response behaviors, such as wearing masks and reducing outings. These measures, by blocking transmission routes, have to some extent curbed the spread of the epidemic. Therefore, the epidemic transmission process and the spread of related public awareness are essentially interactive and co-evolving. How to model and analyze the co-evolutionary transmission process of awareness and epidemics has become a long-term focus of academic research.

[0003] To explore the coupled propagation mechanism between epidemics and related consciousness, scholars generally employ a two-layer network structure to characterize this co-evolution process. In these models, the upper network is a virtual contact layer, where nodes represent real-world individuals and edges represent information exchange between individuals; the lower network is a physical contact layer, where nodes are identical to those in the upper network and edges represent physical contact between individuals. For example, Granell et al. first proposed a consciousness-epidemic co-evolution propagation model on a two-layer network, successfully describing the coupled propagation mechanism of epidemics and related consciousness in the real world. Subsequently, researchers in this field began to further improve and refine the model by incorporating various real-world factors.

[0004] However, many existing studies have neglected the crucial context of social media, failing to fully reflect its important role in the diffusion of awareness and the spread of the epidemic. In fact, social media platforms, through spontaneous user dissemination, social network expansion, and real-time interactive communication, can rapidly amplify the reach and influence of epidemic prevention information, effectively raising public awareness and prompting proactive protective measures. This interaction between information dissemination and behavioral change is particularly critical in today's highly interconnected information environment. Therefore, incorporating social media into the study of the co-evolution of awareness and epidemic transmission has significant practical implications. Consequently, in the current social media environment, there is a need for a computational experimental method and analytical tool that can dynamically capture social media information flow and how to reshape public risk perception, thereby influencing disease transmission pathways. Summary of the Invention

[0005] The purpose of this invention is to overcome the limitations of existing technologies and provide a co-evolutionary propagation analysis method for consciousness-epidemics that takes into account the influence of social media.

[0006] The objective of this invention is achieved through the following technical solution: This invention provides a method for analyzing the co-evolution of awareness and epidemic spread by integrating the influence of self-media. The analysis method includes the following steps: S1, crawling and collecting data from social media platforms and disease prevention and control data, and cleaning and preprocessing the acquired data; S2, based on the social media data and disease prevention and control data integrated in step S1, constructing a two-layer coupled network model of awareness-epidemic co-evolution spread in the self-media environment; S3, using the micro-Markov chain method to analyze the dynamic evolution process of the coupled network model obtained in step S2, establishing the corresponding state transition equation set, and further deriving the critical threshold for epidemic spread under steady-state conditions; S4, based on the analysis results of step S3, quantitatively analyzing the synergistic influence mechanism of self-media propaganda and guidance and individual awareness awakening on epidemic outbreaks.

[0007] Further, step S1 involves crawling and collecting data from social media platforms and disease prevention and control data, and cleaning and preprocessing the acquired data. Specifically, this includes the following steps: S11. Constructing a social media information database for social media platforms and simultaneously collecting key epidemiological data from official health institution data interfaces or publishing platforms; S12. When collecting social media data, after selecting a target social media platform, adding the first user among its reposting users as the starting node to the crawling queue, then extracting users sequentially from the queue, crawling their publicly available personal information and original or reposted Weibo content, while simultaneously tracking the data of the next reposting user, iterating the above process until the target social media data range is covered; S13. Subsequently, using customized data interfaces or webpage structure parsing technology, obtaining key data on epidemic incidence rates, transmission routes, and prevention and control measures from professional medical databases and official health institution websites; S14. Finally, removing noise and duplicate records through data cleaning, and integrating social media and official data into a unified standard format to ensure data quality and support epidemiological analysis.

[0008] Specifically, in S2, a two-layer coupled network model of consciousness-epidemic co-evolution propagation in a self-media environment is constructed, including an upper-layer virtual contact network and a lower-layer physical contact network. The upper-layer virtual contact network is used to describe the diffusion of consciousness, and the lower-layer physical contact network is used to simulate the spread of infectious diseases. Specifically, a coupled network model containing N nodes is considered, whose topology is determined by the adjacency matrix of the upper-layer virtual contact network. Adjacency matrix of lower-level physical contact network The common characteristics described are the process of consciousness diffusion and epidemic spread.

[0009] Specifically, the upper-level virtual contact network is a consciousness diffusion dynamic process constructed based on an unknown-perception-unknown framework, and the network nodes are in a ratio of [missing information]. The ratio of ordinary individuals to self-media nodes on the network The self-media nodes are composed of states including an unknown state U or a perceived state A; specifically, ordinary nodes in state A spread their consciousness to their neighbors with a consciousness diffusion probability λ; self-media nodes in state A, on the other hand, spread their consciousness to their neighbors through enhancement factors. by Consciousness diffuses with probability; simultaneously, all A-state nodes forget their consciousness and return to the U-state with a probability of δ; besides the unknown-perception-unknown UAU process, U-state nodes can also transform into the A-state through two pathways: first, infected U-state nodes awaken their consciousness with a probability of σ due to an infection event; second, U-state nodes that have not completed consciousness interaction with their neighbors... The probability of sensing the media environment and transitioning to state A, where The media attention generated by the pandemic represents the positive effect of self-media on the dissemination of awareness; The proportion of A-state self-media nodes, a baseline parameter. Used to quantify the basic influence of the media environment.

[0010] Specifically, the lower-level physical contact network constructs the infectious disease transmission dynamics based on a susceptible-infected-susceptible framework, and its node states include susceptible state S or infected state I. In this network, I-state nodes transmit the infectious disease to their neighbors with probability β and recover to state S with recovery probability μ. Specifically, when an A-state node interacts with an infected neighbor, its infection probability changes from the base value... Down to , where γ∈ [0,1] is a moderating factor, and this mechanism quantitatively characterizes the inhibitory effect of consciousness arousal on the spread of infectious diseases.

[0011] Furthermore, in step S3, the microscopic Markov chain method is used to analyze the dynamic evolution process of the coupled network model obtained in step S2, and a corresponding set of state transition equations is established. Specifically, each node in the coupled network model has four states: unknown susceptible state US, perceived susceptible state AS, unknown infected state UI, and perceived infected state AI; the probability variables of the four states of node i at time t are defined as follows: , , and Then, the probability of the composite state is derived: , , and Furthermore, the probability that node i is in state U at time t and has not received any consciousness information through any neighboring nodes is: ;in, This represents an element in the adjacency matrix, with a value of 1 or 0, indicating whether there is a connection between node i and node j. This represents the index of a neighboring node, used in the formula to iterate through all neighbors of node i. This indicates the proportion of self-media nodes in the network; The parameter representing the rate of consciousness propagation in the perception state, relative to the base propagation rate λ, reflects the enhanced effect of consciousness information propagation when a node perceives; the probabilities that node i is in the unknown susceptible state US and the perceived susceptible state AS and is not infected at time t are respectively: and ;in, This represents the connection strength or infection probability weight between nodes i and j; This represents the probability of an epidemic spreading under unknown conditions; This represents the probability of epidemic transmission under the perceived state. Then, a state transition probability tree for nodes in a two-layer coupled network is constructed, and all possible state transition paths within a unit time step are derived. Based on this probability tree, the following set of state transition equations is derived using the microscopic Markov chain method: ; ; ; And satisfy: ;in, This represents the probability that node i did not receive any conscious information from its neighbors through the social network at time t; δ represents the media coverage rate, indicating the probability that node i directly obtains consciousness information through global media channels; δ represents the consciousness decay rate, indicating the probability that an individual in the perceived state forgets protective awareness and reverts to an unknown state; μ represents the recovery rate, the probability that an infected person recovers to health. This represents the probability that an individual in a state of awareness, even if infected, will not show symptoms or be detected due to protective measures.

[0012] Furthermore, the derivation of the critical threshold for epidemic transmission under steady-state conditions in S3 specifically involves: after deriving the state transition equation based on the microscopic Markov chain method, calculating the critical threshold for an epidemic outbreak. In the co-evolution dynamics process, when the infection rate exceeds The epidemic will either break out or subside; therefore, it can be determined by the proportion of infected nodes in the system. The epidemic threshold is given, and its calculation formula is as follows: Where N represents the total number of nodes in the network; This represents the individual infection probability at node i; This represents the probability that node i is in an "unknown infection" state; Let represent the probability that node i is in the "perceived infection" state. Starting from the steady-state state transition equation, by adding the unknown infection state transition probability equation to the perceived infection state transition probability equation, we obtain the expression for the infection node probability: When the probability of infection approaches zero, that is... ,right and Approximation is performed; based on the unknown state immunity probability equation and the perceived state immunity probability equation, we obtain: ; where the definition for: Substituting the first-order approximation formulas for the probability of immunity in the unknown state and the probability of immunity in the perceived state into the expression for the probability of infected nodes, the equation for the proportion of infected nodes is simplified to: Among them, in Under the given conditions, derive and Specifically, neglecting in steady state The terms are derived from the probability equations for transitions between unknown and perceived susceptible states: These two equations correspond to a UAU process decoupled from the SIS process, satisfying... Substituting this into the steady-state approximation equation for the node infection probability, we get: Written in matrix form: ;in These are the elements of the identity matrix. Define matrix H with the following elements: The matrix form of the nontrivial solution to the steady-state approximation equation for the node infection probability corresponds to the eigenvector of H, whose eigenvalues ​​satisfy... Therefore, the critical threshold for an epidemic outbreak is given by the largest eigenvalue of H: ;in This represents the largest eigenvalue of matrix H.

[0013] Furthermore, the quantitative analysis in S4 of the synergistic impact mechanism of self-media propaganda and guidance and individual awareness awakening on the outbreak of the epidemic specifically includes: systematically studying the proportion of self-media nodes through the method of controlling variables. Awareness diffusion enhancement factor α, media attention to the epidemic The study investigates the regulatory mechanisms of four key parameters—the probability of individual consciousness arousal (σ)—on the spread and evolution of epidemics. Specifically, based on system steady-state conditions, other parameters are fixed while individual parameters are adjusted to analyze the proportion of infected nodes as each parameter changes. Dynamic response and epidemic threshold The corresponding changing trends of these parameters were studied to reveal the quantitative impact mechanism of the four types of parameters on the epidemic transmission process.

[0014] Compared with existing technologies, this invention offers the following advantages: Most current studies on the co-evolutionary propagation dynamics of awareness and epidemics neglect the crucial background factor of social media, failing to fully reflect its important role in awareness diffusion and epidemic spread. In fact, social media platforms, through spontaneous user dissemination, social network diffusion, and real-time interactive communication, can rapidly amplify the coverage and influence of epidemic prevention information, effectively enhancing public awareness and prompting proactive protective measures. To address this challenge, this invention proposes a co-evolutionary propagation analysis method for awareness and epidemics that considers the influence of social media. By constructing a two-layer coupled network model encompassing awareness diffusion and epidemic spread, it reveals the synergistic inhibitory mechanism of social media propaganda and individual awareness on epidemic outbreaks, providing a quantitative analysis tool for studying the propagation patterns and precise intervention control of sudden epidemics in the social media environment. Attached Figure Description

[0015] To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be described in detail below with reference to the accompanying drawings, wherein: Figure 1 is a schematic diagram of the consciousness-epidemic co-evolution propagation analysis framework considering the influence of self-media in the present invention; Figure 2 is a schematic diagram of data collection and preprocessing in an embodiment of the present invention; Figure 3 is a diagram of the consciousness-epidemic co-evolution propagation model in the context of self-media in an embodiment of the present invention; Figure 4 is a schematic diagram of the state transition probability tree in an embodiment of the present invention; Figure 5 is a schematic diagram of the quantitative analysis framework in an embodiment of the present invention; Figure 6 is a comparison diagram of the microscopic Markov chain method and Monte Carlo simulation results in an embodiment of the present invention; wherein, (a) Figure shows the change of the steady-state proportion of individuals in the perceptual state with the infection probability; (b) Figure shows the change of the steady-state proportion of individuals in the infected state with the infection probability; Figure 7 is a diagram showing the influence of the proportion of self-media nodes on the dynamics of infectious disease transmission in an embodiment of the present invention; wherein, (a) Figure shows the influence of different proportions of self-media nodes (values ​​of 0, 0.1, 0.3, 0.6, ... Figure 8 shows the change in the steady-state proportion of infected individuals with the infection probability at a fixed infection probability of 0.9; (b) Figure 8 shows the change in the steady-state proportion of infected individuals with the proportion of self-media nodes when the infection probability is fixed at 0.2; Figure 8 is a diagram showing the influence of the self-media awareness diffusion enhancement factor on the dynamics of infectious disease transmission in the embodiments of the present invention; wherein, (a) Figure 8 shows the change in the steady-state proportion of infected individuals with the self-media awareness diffusion enhancement factor (values ​​are 1, 5, 10, 20, ...) at different self-media awareness diffusion enhancement factors ... Figure 30 shows the change in the steady-state proportion of infected individuals with the infection probability; (b) shows the change in the steady-state proportion of infected individuals with the social media awareness diffusion enhancement factor when the infection probability is fixed at 0.3; Figure 9 is a diagram showing the influence of epidemic media popularity on the transmission dynamics of infectious diseases in the embodiments of the present invention; wherein, (a) shows the relationship between the steady-state proportion of infected individuals and the infection probability when the regulation factor is 1; (b) shows the relationship between the steady-state proportion of infected individuals and the infection probability when the regulation factor is 0.5; (c) shows the relationship between the steady-state proportion of infected individuals and the infection probability when the regulation factor is 0.25; (d) shows the relationship between the steady-state proportion of infected individuals and the infection probability when the regulation factor is 0. Detailed Implementation

[0016] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. The components of the embodiments of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the detailed description of the embodiments of this application provided below with reference to the accompanying drawings is not intended to limit the scope of protection of the claimed application, but merely represents selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application. The present invention will be further described below with reference to the accompanying drawings.

[0017] This invention proposes a co-evolutionary propagation analysis method for awareness and epidemics that integrates the influence of social media, addressing the limitations of existing epidemic models that do not fully consider the key impacts of social media on public awareness and epidemic spread. First, a two-layer coupled network model of awareness-epidemic co-evolutionary propagation in the context of social media is constructed by integrating social media and disease control data. The upper-layer virtual contact network achieves awareness diffusion through the interaction between ordinary individuals and social media nodes, while the lower-layer physical contact network nodes correspond one-to-one with the upper layer, simulating the transmission process of infectious diseases. Based on this, a Microscopic Markov-Chain Approach (MMCA) is proposed to analyze the model's dynamics. By establishing a set of state transition equations and deriving the critical threshold under steady-state conditions, the impact of social media propaganda and individual awareness on epidemic outbreaks is quantitatively analyzed. This invention provides a quantitative analysis tool for studying the propagation patterns and precise intervention and control of sudden epidemics in the social media environment.

[0018] As shown in Figure 1, the specific steps include: Step 1: Constructing a social media information database for social media platforms (taking Weibo as an example), while simultaneously collecting key epidemiological data from authoritative sources (data interfaces (such as APIs) or publishing platforms of official health institutions). As shown in Figure 2, firstly, publicly available social media data is collected. Specifically, a target Weibo post is selected, and the first user among its reposting users is added to the queue to be crawled; then, users are extracted sequentially from the queue, their personal information and original / reposted Weibo content are crawled, and the data of the next reposting user is tracked, repeating the above process until the required social media data range is covered. Next, through customized data interfaces or webpage structure parsing, data is obtained from authoritative sources such as official health institution websites and professional medical databases to collect key data such as epidemic incidence rates, transmission paths, and prevention and control measures. Finally, the collected data undergoes standardization processing such as cleaning and integration; the standardization processing specifically involves: removing noise and duplicate records through data cleaning, and integrating social media and official data into a unified standard format to ensure data quality and support for epidemiological analysis.

[0019] Step 2: Based on the social media data and disease control data integrated in Step 1, construct a two-layer coupled network model of awareness-epidemic co-evolution propagation in the self-media environment. As shown in Figure 3, consider a coupled network model containing N nodes, whose topology consists of an upper-layer virtual contact network adjacency matrix. Adjacency matrix of lower-level physical contact network Common characteristics describe the diffusion of consciousness and the spread of the epidemic.

[0020] The virtual contact network constructs a dynamic process of consciousness diffusion based on an unknown-perception-unknown framework. The network nodes are arranged in a ratio of... The ratio of ordinary individuals to self-media nodes on the network The system consists of self-media nodes, with states including unknown (U) or perceived (A). Specifically, ordinary nodes in state A spread their consciousness to their neighbors with a consciousness diffusion probability λ; self-media nodes in state A spread their consciousness to their neighbors with an enhancement factor α ≥ 1. Consciousness diffuses with probability; simultaneously, all A-state nodes forget their consciousness and revert to U-state with probability δ. Besides the unknown-perception-unknown UAU process, U-state nodes can also transform into A-state nodes through two pathways: first, infected U-state nodes awaken consciousness with probability σ due to an infection event; second, U-state nodes that have not completed consciousness interaction with their neighbors... It changes to state A based on probability perception of the media environment, where... The media attention surrounding the pandemic reflects the amplifying effect of social media on the dissemination of awareness. The proportion of A-state self-media nodes, a baseline parameter. Used to quantify the fundamental influence of the media environment; the Susceptible-Infected-Susceptible (SIS) framework of physical contact networks constructs the dynamic process of infectious disease transmission, where node states include susceptible (S) or infected (I); in this network, I-state nodes spread the infectious disease to their neighbors with probability β and recover to the S state with recovery probability μ; specifically, when an A-state node interacts with an infected neighbor, its infection probability changes from the base value. Down to Where γ ∈ [0, 1] is a moderating factor, this mechanism quantitatively characterizes the inhibitory effect of consciousness arousal on the spread of infectious diseases; Step 3: The microscopic Markov chain method is proposed to analyze the dynamic process of the model. Each node in the model has four states: unknown susceptible state (US), perceived susceptible state (AS), unknown infected state (UI), and perceived infected state (AI). The probability variables of the four states of node i at time t are defined as follows: , , and Therefore, the probability of the composite state is derived as follows: , , and Furthermore, the probability that node i is in state U at time t and has not received any consciousness information through any neighboring nodes is: in, This represents an element in the adjacency matrix, with a value of 1 or 0, indicating whether there is a connection between node i and node j. This represents the index of a neighboring node, used in the formula to iterate through all neighbors of node i. This indicates the proportion of self-media nodes in the network; This parameter represents the rate of consciousness propagation under the perception state. It is relative to the basic propagation rate λ and reflects the enhanced effect of consciousness information propagation when nodes perceive.

[0021] The probabilities that node i is in both the Unknown Susceptible State (US) and the Perceived Susceptible State (AS) and is not infected at time t are respectively and in, This represents the connection strength or infection probability weight between nodes i and j; This represents the probability of an epidemic spreading under unknown conditions; This represents the probability of an epidemic spreading under perceived conditions.

[0022] Figure 4 illustrates the state transition probability tree of a node in a two-layer coupled network, describing all possible state transition paths within a unit time step. Based on this probability tree, the following set of state transition equations is derived using the MMCA method: and satisfy in, This represents the probability that node i did not receive any conscious information from its neighbors through the social network at time t; δ represents the media coverage rate, indicating the probability that node i directly obtains consciousness information through global media channels; δ represents the consciousness decay rate, indicating the probability that an individual in the perceived state forgets protective awareness and reverts to an unknown state; μ represents the recovery rate, the probability that an infected person recovers to health. This indicates the probability that an individual in a state of awareness may not show symptoms or be undetected even if infected, due to protective measures.

[0023] Step 4: Establish a system of state transition equations and derive the critical threshold under steady-state conditions. After deriving the state transition equations based on the Microscopic Markov Chain Method (MMCA), the critical threshold for an epidemic outbreak can be calculated. In the co-evolution dynamics process, when the infection rate exceeds... An epidemic will either break out or subside. Therefore, the epidemic threshold can be determined by the proportion of infected nodes in the system. The calculation formula is given as follows: Where N represents the total number of nodes in the network; This represents the individual infection probability at node i; This represents the probability that node i is in an "unknown infection" state; Let i represent the probability that node i is in the "perceived infection" state. Starting from the steady-state state transition equation, by adding equations (5) and (7), a more concise expression for the probability of infected nodes can be obtained: When the probability of infection is close to zero (i.e.) ), can be used for and Approximation is performed. Based on equations (2) and (3), we obtain: Where the definition for: Substituting equations (11) and (12) into the expression for the probability of infected nodes, equation (10) can be simplified to: Among them, Under the given conditions, it can be deduced that and Similarly, neglecting in steady state The term can be obtained from equations (4) and (6): These two equations correspond to a UAU process decoupled from the SIS process, satisfying Substituting it into equation (14), we get: It can be written in matrix form: in These are the elements of the identity matrix. Define matrix H with the following elements: Then the nontrivial solution of equation (19) corresponds to the eigenvector of H, and its eigenvalues ​​satisfy... Therefore, the critical threshold for an epidemic outbreak is given by the largest eigenvalue of H: in This represents the largest eigenvalue of matrix H.

[0024] Step 5: Quantitatively analyze the synergistic impact mechanism of self-media propaganda and guidance and individual awareness awakening on the epidemic outbreak. As shown in Figure 5, the self-media parameters (node ​​proportions) are systematically studied using the controlled variable method. Awareness diffusion enhancement factor α, media attention to the epidemic The study investigated the synergistic regulatory relationship between the probability of individual consciousness arousal (σ) and the evolution of epidemic transmission. Specifically, based on system steady-state conditions and keeping other parameters constant, four key parameters were sequentially adjusted to different values. Then, the proportion of infected nodes was recorded in real time. The numerical trend and the epidemic threshold were calculated simultaneously. The corresponding changing trends of the four types of parameters are then analyzed. Finally, by combining the dynamic responses and threshold change patterns of the four types of parameters, the synergistic influence mechanism on the spread and evolution of the epidemic is quantitatively analyzed.

[0025] Example 1: Based on the above analysis method, this example uses a two-layer coupled network with 1000 nodes for numerical simulation to verify the effectiveness of the method and quantitatively analyze the impact of key parameters on the propagation process. In the experimental setup, the physical contact layer uses a scale-free network structure with a degree distribution index of 2.5 and an average node degree of 6 to simulate the non-uniform interpersonal contact pattern in real society. The virtual contact layer introduces self-media nodes and increases the edge density, with an average degree of 10 to reflect the high-frequency interaction characteristics of online social networks. First, the accuracy of the Microscopic Markov Chain Method (MMCA) is verified by using the steady-state infection density calculated iteratively from MMCA. and the theoretical value of consciousness density The results were compared with those from Monte Carlo (MC) stochastic simulations. As shown in Figure 6, the fitting curves of the two methods under different infection probabilities highly overlap, confirming that the state transition equations derived in this invention can accurately describe the dynamic evolution of the system.

[0026] Based on this simulation environment, step S4 is executed to quantitatively analyze the proportion of self-media nodes. The regulatory role of social media in the spread of epidemics. With other parameters fixed, the proportion of social media nodes in the network was gradually increased, and the steady-state infection density was observed. The experimental results, as shown in Figure 7, indicate that as the proportion of self-media nodes increases from 0 to 0.3, the steady-state infection density exhibits a sharp downward trend, demonstrating that the intervention of self-media can rapidly expand information coverage and effectively suppress the spread of the epidemic. However, when the proportion continues to increase and exceeds 0.5, the downward trend in infection density tends to level off, showing a clear saturation effect. This finding reveals that in actual intervention, blindly increasing the number of self-media accounts does not guarantee sustained high returns; rather, there exists an optimal deployment ratio range based on cost-effectiveness.

[0027] Subsequently, the factors enhancing the spread of self-media awareness were analyzed. Impact on propagation dynamics. The enhancement factor characterizes the multiple advantage of self-media nodes in information dissemination capacity compared to ordinary nodes. Simulation results shown in Figure 8 indicate that as the enhancement factor value increases, the steady-state infection density in the system decreases significantly. This is attributed to the fact that high-influence self-media accelerates the penetration of risk awareness in the network. Similarly, this parameter also exhibits diminishing marginal returns. Once the enhancement factor reaches a certain threshold (e.g., 20), further increasing its propagation intensity has negligible effect on suppressing the epidemic, indicating that there is an upper limit to the effectiveness of relying solely on excessive exposure from top influencers, and that optimal control can only be achieved by combining it with widespread grassroots dissemination.

[0028] Further examine the media coverage of the epidemic The synergistic mechanism between individual protection effectiveness and the overall media popularity parameter was introduced into the experiment. This study simulates the impact of a non-contact, global information environment. Analysis of Figure 9 shows that the effectiveness of media attention in suppressing the epidemic is highly dependent on the efficiency of preventative measures after awareness is raised. When individuals take highly effective protective measures after becoming aware, increasing media attention can significantly reduce the scale of infection; conversely, if preventative measures after awareness are ineffective, simply increasing media attention cannot change the course of the epidemic. This result quantifies the coupling relationship between "publicity and guidance" and "behavioral implementation," indicating that media bombardment must be combined with effective personal protective measures to have a substantial effect.

[0029] Finally, to verify the applicability of this analysis method in real-world scenarios, this embodiment selects the topology parameters of the real network in the embodiment, such as the topology data of real social networks like Facebook and LastfmAsia shown in Table 1, to construct the upper-layer virtual network and generate the corresponding lower-layer physical network for full-process analysis.

[0030] Table 1 Under real-world topology, the key conclusions regarding the proportion of self-media, enhancement factors, and media popularity remain valid, and improvements in all self-media-related parameters effectively raise the critical threshold for epidemic outbreaks. This means that by conducting pre-analysis using the method of this invention, it is possible to determine how to combine and configure self-media resources (such as account density, dissemination intensity, and overall network popularity) under specific network structures, thereby significantly improving the overall defense threshold of the system in the early stages of an outbreak, making it more difficult for infectious diseases to spread on a large scale among the population. This embodiment fully demonstrates that this method can provide reliable data support and theoretical basis for public health departments to formulate precise intervention strategies based on social media.

[0031] This invention designs a co-evolutionary propagation analysis method for awareness-epidemic transmission that considers the influence of social media. Firstly, addressing the limitation of existing epidemic models that do not fully consider the key impact of social media on public awareness and epidemic spread, a two-layer coupled network model of awareness-epidemic co-evolutionary transmission in the context of social media is constructed by integrating social media and disease control data. In this model, the upper-layer virtual contact network achieves awareness diffusion through the interaction between ordinary individuals and social media nodes, while the lower-layer physical contact network nodes correspond one-to-one with the upper layer, simulating the transmission process of infectious diseases. Based on this, a micro-Markov chain method is proposed to analyze the dynamic process of the model. By establishing a set of state transition equations and deriving the critical threshold under steady-state conditions, the impact of social media propaganda and individual awareness on epidemic outbreaks is quantitatively analyzed. This invention provides a quantitative analysis tool for studying the transmission patterns of sudden epidemics in the social media environment and for precise intervention and control.

[0032] The above description is merely a preferred embodiment of the present invention. It should be understood that the present invention is not limited to the forms disclosed herein and should not be construed as excluding other embodiments. It can be used in various other combinations, modifications, and improvements, and can be altered within the scope of the concept described herein through the above teachings or related technologies or knowledge. Modifications and variations made by those skilled in the art that do not depart from the spirit and scope of the present invention should be within the protection scope of the appended claims.

Claims

1. A method for analyzing the co-evolution of awareness and epidemics by integrating the influence of self-media, characterized in that, The analytical method includes the following steps: S1, crawling and collecting data from social media platforms and disease prevention and control data, and cleaning and preprocessing the acquired data; S2, based on the social media data and disease prevention and control data integrated in step S1, constructing a two-layer coupled network model of awareness-epidemic co-evolution and transmission in the self-media environment; S3, using the micro-Markov chain method to analyze the dynamic evolution process of the coupled network model obtained in step S2, establishing the corresponding state transition equation set, and further deriving the critical threshold of epidemic transmission under steady-state conditions; S4, based on the analysis results of step S3, quantitatively analyzing the synergistic influence mechanism of self-media propaganda and guidance and individual awareness awakening on epidemic outbreaks.

2. The method for analyzing the co-evolution of awareness and epidemic spread by integrating the influence of self-media, as described in claim 1, is characterized in that... Step S1 involves crawling and collecting data from social media platforms and disease prevention and control data. The acquired data is then cleaned and preprocessed, specifically including the following steps: S11. Constructing a social media information database for the social media platforms and simultaneously collecting key epidemiological data from official health institution data interfaces or publishing platforms; S12. When collecting social media data, after selecting a target social media platform, the first user among its reposting users is added to the crawling queue as the starting node. Users are then extracted sequentially from the queue, and their publicly available personal information and original or reposted Weibo content are crawled. Simultaneously, the data of the next reposting user is tracked, iterating the above process until the target social media data range is covered; S13. Subsequently, key data on epidemic incidence rates, transmission routes, and prevention and control measures are obtained from professional medical databases and official health institution websites through customized data interfaces or webpage structure parsing technology; S14. Finally, data cleaning removes noise and duplicate records, and social media and official data are integrated into a unified standard format to ensure data quality and support epidemiological analysis.

3. The method for analyzing the co-evolution of awareness and epidemic spread by integrating the influence of self-media, as described in claim 1, is characterized in that... S2 constructs a two-layer coupled network model for the co-evolution and spread of consciousness and epidemics in a self-media environment, including an upper-layer virtual contact network and a lower-layer physical contact network. The upper-layer virtual contact network is used to describe the diffusion of consciousness, and the lower-layer physical contact network is used to simulate the spread of infectious diseases. Specifically, a coupled network model containing N nodes is considered, whose topology is determined by the adjacency matrix of the upper-layer virtual contact network. Adjacency matrix of lower-level physical contact network The common characteristics described are the process of consciousness diffusion and epidemic spread.

4. The method for analyzing the co-evolution of awareness and epidemic spread by integrating the influence of self-media, as described in claim 3, is characterized in that... The upper-level virtual contact network is a consciousness diffusion dynamic process constructed based on an unknown-perception-unknown framework. The network nodes are in a ratio of... The ratio of ordinary individuals to self-media nodes on the network The self-media nodes are composed of states including an unknown state U or a perceived state A; specifically, ordinary nodes in state A spread their consciousness to their neighbors with a consciousness diffusion probability λ; self-media nodes in state A, on the other hand, spread their consciousness to their neighbors through enhancement factors. by Consciousness diffuses with probability; simultaneously, all A-state nodes forget their consciousness and return to the U-state with a probability of δ; besides the unknown-perception-unknown UAU process, U-state nodes can also transform into the A-state through two pathways: first, infected U-state nodes awaken their consciousness with a probability of σ due to an infection event; second, U-state nodes that have not completed consciousness interaction with their neighbors... The probability of sensing the media environment and transitioning to state A, where The media attention generated by the pandemic represents the positive effect of self-media on the dissemination of awareness; The proportion of A-state self-media nodes, benchmark parameter Used to quantify the basic influence of the media environment.

5. The method for analyzing the co-evolution of awareness and epidemic spread by integrating the influence of self-media, as described in claim 3, is characterized in that... The underlying physical contact network is based on a susceptible-infected-susceptible framework to construct the dynamics of infectious disease transmission. Its node states include susceptible state S or infected state I. In this network, I-state nodes transmit the infectious disease to their neighbors with probability β and recover to state S with recovery probability μ. Specifically, when an A-state node interacts with an infected neighbor, its infection probability changes from the base value... Down to , where γ∈ [0,1] is a moderating factor, and this mechanism quantitatively characterizes the inhibitory effect of consciousness arousal on the spread of infectious diseases.

6. The method for analyzing the co-evolution of awareness and epidemic spread by integrating the influence of self-media, as described in claim 1, is characterized in that... In step S3, the microscopic Markov chain method is used to analyze the dynamic evolution of the coupled network model obtained in step S2, and a corresponding set of state transition equations is established. Specifically, each node in the coupled network model has four states: unknown susceptible state US, perceived susceptible state AS, unknown infected state UI, and perceived infected state AI; the probability variables of the four states of node i at time t are defined as follows: 、 、 and Then, the probability of the composite state is derived: 、 、 and Furthermore, the probability that node i is in state U at time t and has not received any consciousness information through any neighboring nodes is: ;in, This represents an element in the adjacency matrix, with a value of 1 or 0, indicating whether there is a connection between node i and node j. This represents the index of a neighboring node, used in the formula to iterate through all neighbors of node i. This indicates the proportion of self-media nodes in the network; The parameter representing the rate of consciousness propagation in the perception state, relative to the base propagation rate λ, reflects the enhanced effect of consciousness information propagation when a node perceives; the probabilities that node i is in the unknown susceptible state US and the perceived susceptible state AS and is not infected at time t are respectively: and ;in, This represents the connection strength or infection probability weight between nodes i and j; This represents the probability of an epidemic spreading under unknown conditions; This represents the probability of epidemic transmission under the perceived state. Then, a state transition probability tree for nodes in a two-layer coupled network is constructed, and all possible state transition paths within a unit time step are derived. Based on this probability tree, the following set of state transition equations is derived using the microscopic Markov chain method: ； ； ； And satisfy: ;in, This represents the probability that node i did not receive any conscious information from its neighbors through the social network at time t; δ represents the media coverage rate, indicating the probability that node i directly obtains consciousness information through global media channels; δ represents the consciousness decay rate, indicating the probability that an individual in the perceived state forgets protective awareness and reverts to an unknown state; μ represents the recovery rate, the probability that an infected person recovers to health. This represents the probability that an individual in a state of awareness, even if infected, will not show symptoms or be detected due to protective measures.

7. The method for analyzing the co-evolution of awareness and epidemic spread by integrating the influence of self-media, as described in claim 1, is characterized in that... The derivation of the critical threshold for epidemic transmission under steady-state conditions in S3 specifically involves: after deriving the state transition equation based on the microscopic Markov chain method, calculating the critical threshold for an epidemic outbreak. In the co-evolution dynamics process, when the infection rate exceeds The epidemic will either break out or subside; therefore, it can be determined by the proportion of infected nodes in the system. The epidemic threshold is given, and its calculation formula is as follows: Where N represents the total number of nodes in the network; This represents the individual infection probability at node i; This represents the probability that node i is in an "unknown infection" state; Let represent the probability that node i is in the "perceived infection" state. Starting from the steady-state state transition equation, by adding the unknown infection state transition probability equation to the perceived infection state transition probability equation, we obtain the expression for the infection node probability: When the probability of infection approaches zero, that is... ,right and Approximation is performed; based on the unknown state immunity probability equation and the perceived state immunity probability equation, we obtain: ; where the definition for: Substituting the first-order approximation formulas for the probability of immunity in the unknown state and the probability of immunity in the perceived state into the expression for the probability of infected nodes, the equation for the proportion of infected nodes is simplified to: Among them, in Under the given conditions, derive and Specifically, neglecting in steady state The terms are derived from the probability equations for transitions between unknown and perceived susceptible states: These two equations correspond to a UAU process decoupled from the SIS process, satisfying... Substituting this into the steady-state approximation equation for the node infection probability, we get: Written in matrix form: ;in These are the elements of the identity matrix. Define matrix H with the following elements: The matrix form of the nontrivial solution to the steady-state approximation equation for the node infection probability corresponds to the eigenvector of H, whose eigenvalues ​​satisfy... Therefore, the critical threshold for an epidemic outbreak is given by the largest eigenvalue of H: ;in This represents the largest eigenvalue of matrix H.

8. The method for analyzing the co-evolution of awareness and epidemic spread by integrating the influence of self-media, as described in claim 1, is characterized in that... The quantitative analysis in S4 of the mechanism by which self-media propaganda and guidance and individual awareness awakening affect the outbreak of an epidemic specifically includes: systematically studying the proportion of self-media nodes using the controlled variable method. Awareness diffusion enhancement factor α, media attention to the epidemic The study investigates the regulatory mechanisms of four key parameters—the probability of individual consciousness arousal (σ)—on the spread and evolution of epidemics. Specifically, based on system steady-state conditions, other parameters are fixed while individual parameters are adjusted to analyze the proportion of infected nodes as each parameter changes. Dynamic response and epidemic threshold The corresponding changing trends of these parameters were studied to reveal the quantitative impact mechanism of the four types of parameters on the epidemic transmission process.

Images