Applying mining fuzzy sequential patterns technique to predict the leadership in social networks

Abstract

Social Network has become a very popular communication tool among Internet users who connect to each other by one or more relations. Millions of users are sharing opinions and experiences on different aspects of life every day via the social network community. Comments from the members of social network can be influential and have an impact on trust among members within the social group. The Discovery of Influential Behavior Pattern of Members in Social Networks brings challenge to workers in this research field. Understanding social networks requires analysis of structural relations between the users and the patterns of interaction among users. This paper thus focuses on 2 folds: First, we defined many factors for leadership in social network and showed that, those at individuals who are central to social networks serve as the opinion leaders. Second, we proposed a fuzzy data-mining algorithm to find association rules for analyzing the posting messages into quantitative values and discovered interesting sequential patterns among them. The sequential patterns are very important for real-world applications since the patterns mined out exhibit the sequential quantitative regularity in databases and can provide useful information to applications.