| ObjectiveTaking public health emergencies as the classification scenario,analyze the various factors that affect the classification of netizens in public health emergencies,comprehensively consider the characteristics of netizens and their relationship with the incident from multiple aspects,and comprehensively utilize the characteristics of various types of netizens.Analyze the characteristics of netizens from multiple dimensions,and build a multi-dimensional and fine-grained classification model of netizens in this scenario.MethodBy summarizing the relevant research on the classification of netizens at home and abroad,combining the characteristics of netizens,such as netizens' own personality characteristics,interest preferences,activity,etc.,and specific classification scenarios,integrating the relationship between netizens and the event,and the role of netizens in the event,according to Related theories have determined the four dimensions of netizen classification,online personality label,netizen influence,event relevance,and topic attention.Use Word2 vec to expand the simplified Chinese version of the LIWC dictionary to build a personality dictionary,based on blog texts published by netizens,calculate network personality scores based on the personality dictionary,analyze network personality labels;determine the three characteristic groups of communication power,activity,and credibility through the analytic hierarchy process And the weights of specific feature indicators,construct the influence formula,calculate the influence of netizens;use Python tools to write programs to match the characteristics of the correlation between netizens and events,and calculate the correlation degree of events;extract netizens through TF-IDF and LDA theme models The feature words of blog texts and event-related texts are calculated using vector similarity to analyze the interest of netizens on the topic of the event.Furthermore,it conducts a comprehensive analysis based on the four dimensions and the methods in each dimension to construct an Internet user classification model-the ICN model,and selects a target event to present the results of the multi-dimensional classification of Internet users.ResultsThrough the constructed netizen classification model,netizens are divided into eight categories.Through a comprehensive analysis of the four dimensions,the degree of event relevance and topic attention jointly determine the relevance of netizens and events.From these two dimensions,the correlation between netizens and events is jointly analyzed,and netizens are classified into high impact from multiple dimensions.Highly relevant and high neuroticism(Influence,Correlation,Neuroticism-High,ICN-H for short)netizens,high-impact high-correlation low-neurotic(IC-H)netizens,high-impact low-correlation high-neurotic(IN-H)netizens,low impact and high correlation High neuroticism(CN-H)netizens,high-impact low-correlated low-neurotic(I-H)netizens,low-impact high-correlation low-neurotic(C-H)netizens,low-impact low-correlated high-neurotic(N-H)netizens,low-impact low-correlation low neuroticism(Influence,Correlation,Neuroticism-Low,ICN-L)netizens.Through the classification model of netizens constructed from four dimensions and specific analysis methods,the Wuhan city closure event is selected to present the classification results of netizens,and according to the characteristics of public health emergencies,the characteristic index of the location of netizens under the event correlation degree is selected to carry out the event.According to the matching calculation of the degree of relevance,122 netizens with a higher degree of relevance were obtained under the Wuhan lockdown event.Use TF-IDF keyword extraction and LDA topic model to mine the interest preference of Weibo netizens and extract the topic feature of the event.By calculating the semantic similarity between the key feature words of netizens and the topic words of public health emergencies,under the Wuhan city closure event 496 netizens with a high degree of attention on topics of interest were obtained.Calculating the influence of netizens through the three characteristic groups of communication power,activity,and credibility and the specific characteristic indicators of each group,446 netizens with high influence were obtained under the Wuhan lockdown.The personality dictionary was obtained by expanding the simplified Chinese version of the LIWC dictionary,using the Pearson coefficient of each category of vocabulary and personality traits in the personality dictionary to calculate the network personality scores.342 netizens with high neuroticism were obtained under the Wuhan lockdown event.Conclusion(1)Comprehensively analyze the internal and external characteristics of netizens,combine their own characteristics,the relationship between netizens and other netizens,and the relationship between netizens and events.From these perspectives,combine the Big Five personality theory,value accumulation theory,interpretation level theory,With reference to group theory,two-level communication theory and related research,interest mining and personality traits are introduced into classification thinking,and four classification dimensions of netizen classification are determined-online personality label,event relevance,topic attention,and netizen influence.This dimension divides netizens into eight categories,and builds a multi-dimensional classification model of netizens—the ICN model.(2)In the context of public health emergencies,construct a multi-dimensional netizen classification model,comprehensively analyze netizens from various dimensions,and can classify netizens in a multi-dimensional and fine-grained manner.The four dimensions can analyze netizens from whether they are more likely to generate negative emotions,the degree of relevance of netizens and events,the degree of interest and attention of netizens to events,and the strength of netizens' influence on the Internet,so that each type of netizen group can be divided.There is a more detailed description of the characteristics. |
PDF Full Text Request