Akademik Veri Yönetim Sistemi
12th International Conference on Machine Learning and Applications (ICMLA), Florida, Amerika Birleşik Devletleri, 4 - 07 Aralık 2013, ss.281-287
Many social network sites have been develop rapidly in recent years. Millions of people are becoming to members of these sites because of facilities provided by these networks. Thousands of people share hundred of comments, news, articles, pictures and videos to support ideas. Recent events have shown that. Users are well and quickly organized for events, demonstrations, boycotts, etc. through social networking sites accounts especially using Twitter accounts. This paper presents an intelligent solution by analysing users' information to clarify if they are supporters or anti-supporters to an event this anaylsis is based on written words or characters on tweets of Twitter accounts. The proposed method presented in this work achieves the task with 90% accuracy within test tweets. Even if the privacy concern is a hot topic in social networks, this sort of solution should provide clear vision to the users and might be used for crime analysing and predicting crime, boycotts, demonstration, events, etc.