Study On The Strength Of Social Relationships Based On Social Media

Since the development of the Internet,a variety of new technologies,new ideas and new applications are emerging,which bring innovations into people's everyday lives.Social Media,which is user-centric and allows users to create content,share content,and discuss content,has become the brightest new star in the current Internet world."Social" is the core concept of Social Media,and its foundation is the social relationships between users.However,the social relationships in Social Media are a little different from those in real life,because currently,they can only indicate that the users are "friends" or "not friends",and they have no "strength" property that is very important in the real social relationships.This paper,taking the strength of social relationships between users as the research objective,uses the data of Social Media users to explore how to automatically distinguish the strength of social relationships,and how to visually present the strength of user's social relationships.Firstly,this paper introduces the research background and the significance of this research.Then,based on the comprehensive review of existing domestic and foreign researches,the research target,research methods,the main research content and the research framework are introduced.In the theoretical foundation section,this paper summarizes the relevant theories from the aspects of social interaction,social networks and social relationships.Then,the factors influencing the strength of social relationships and the classification of the strength of social relationships have been the following focus of our discussion,and finally,we summarize the factors into three categories:similarity factors,time factors and interactivity factors,and according to different levels of the strength of social relationships,we also explicitly classify the social relationships into strong social relationships,weak social relationships and temporary social relationships.In the section of research environments and research tools,this paper selects RenRen as the specific research platform based on the detailed analysis of Social Media,and introduces the research tools that be used in this study,including self-developed data collection tools,data mining tool and data visualization tool.In the classification study of the strength of social relationships,we explore the best classification models in two steps,with the supervised Machine Learning methods and the personal data and interaction data from the users of RenRen.The findings indicate that the classification model based on BayesNet algorithm is proved to be most effective when distinguishing the strong social relationships(the first step),and the classification model based on Logistic Regression algorithm has the best performance when distinguishing the temporary social relationships(the second step).Moreover,with the help of attribute analysis and error analysis,we find that,overall,the interactivity factors have the most prominent distinguish ability for the strength of social relationships,and the common friends number in similarity factors also has a good distinguish ability.But,the distinguish ability of the only time factor(the number of days from the friend's recent visit to the user's home page)has not been excavated.In order to visually present the distribution of the social relationships with different strength,we generate the "Social Relationship Strength Graph" on a web page using the data visualization tool,and we also evaluate the graph,according to the usability evaluation system,by use of the semi-structured interview method.The evaluation results show that the "Social Relationship Strength Graph" can basically reflect the real strength of the social relationships between users and their friends,and the graph can also help users to quickly select their social relationships of different strength,and the comprehensibility of the graph can make the users satisfied too.After the analysis and discussion of the evaluation results and interview answers,we also find many issues worthy of further attention and many shortcomings that need to be improved.This study has some theoretical contributions,including:1)Using the new method and new research data in the Internet environment,expands the study of traditional sociological theories and provides a new theoretical support for the Internet-based study of sociological issues.2)The summary of the factors that affect the strength of social relationship and the classification system of strength of social relationship contribute to enrich the theories of social relationship strength in Sociology of the Internet.The practical contributions of this study could be concluded as:1)Helping Social Media use the classification of strength of social relationships and the "Social Relationship Strength Graph" to develop a more rational and more fine-grained user privacy control function.2)Being conducive to Social Media to implement more effective information filtering functions based on the strength of social relationships.3)Suggesting Social Media to use the strength of social relationships from multi-angle to optimize its services so as to enhance the user experience.