Network Characteristics And Interpersonal Relationship Prediction Of High Social Exclusion Individuals:Based On Social Network Analysis

Social exclusion,as a negative interpersonal experience,generally exists in college students’ interpersonal communication.The study found that social exclusion can cause a series of mental health problems and even anti-social behavior.However,in the past,the research on social exclusion seldom considered specific social situations and lacked the perspective of group research.Social exclusion measures the relationship between groups and individuals,which is limited to the subjective report of individuals and can not fully reflect their social status,and this relationship is situational.College students have a variety of social relationships,and a popular individual in one circle may turn into a marginal figure in another circle.So,is there such a situation for the excluded? In addition,although the excluded person may be an important person in some aspects,the individual who perceives exclusion is at a disadvantage in many aspects.For the negative phenomenon of social exclusion,how to intervene or prevent is particularly critical,so how to strengthen the relationship between individuals and groups through new methods? Due to the limited perspective of previous studies,it is difficult to study these issues.This paper uses social network analysis to supplement it and discuss the above issues.This study includes two parts.Firstly,individuals are placed in the group network for analysis through the group perspective.According to different social situations,nine social networks are used.The concept of node centrality in social network analysis describes the importance of individuals in the network,and then explores the relationship between social exclusion and node centrality in different networks to find out whether individuals with high social exclusion are important people in the network under certain circumstances.Finally,the link prediction technology in social network analysis is used to predict the objects most likely to have relations with individuals in the current situation,providing a feasible auxiliary method for the current social exclusion intervention.In study 1,115 subjects were investigated.They were college students of the same major and grade from a university in Sichuan,belonging to the same group.The social exclusion questionnaire and social network questionnaire were used for data collection,and the network level analysis was carried out to observe the differences and connections of various social situations.Node centrality is generated by the network structure and calculated by programming.There are five types.By grouping social exclusion into high and low groups,we finally found that the high group has the same importance as the low group in some aspects in social advice,seeking help,news sharing,understanding,and positive feelings network.In study 2,nine social networks obtained in study 1 were used to predict links based on ten similarity indicators.AUC,Precision and RS are used to judge the prediction performance.It is found that most similarity indicators have good prediction performance.It proves the feasibility of applying link prediction to interpersonal relationship prediction,and provides a new auxiliary way for the intervention of social exclusion.These results illustrate the importance of considering specific situations and relationships between individuals when studying interpersonal relationships.By combining network science with psychology,this paper makes up for the limitations of previous research to some extent,and the prediction of friendship also expands the perspective of existing intervention research.It has made certain contributions to better grasp the social law of college students.To sum up,this paper draws the following conclusions:1.There are differences and close connections between social situations generated by different social needs.2.Individuals with high social exclusion are not at the edge of the group in all social situations.In some social situations,they have the same importance as individuals with low social exclusion.3.In this paper,when link prediction is applied to interpersonal relationship prediction,CN,jaccard,S?rensen and Salton indicators have achieved good prediction results in different social situations.