DATA MINING APPROACH FOR CLASSIFYING TWITTER’S USERS

Social networks are the most important communication channels in recent years, which popular among the different social groups. These networks affected the ideas and policies of individuals, groups and communities. Every day, millions of tweets on Twitter are being published. These tweets reflect opinions and beliefs of their publishers and affect others as well. Therefore, it is important to analyze these tweets and identify and classify trends of different users. This research aims to classify social network to anomaly groups such as: Terrorist and dissident; by analyzing tweets data on the Twitter; then identify an anonymous user's affiliation to these groups. To address this problem, we first extract a set of features to characterize each group using different data mining techniques and store these features in the database. Text mining, sentiment analysis, and opinion mining techniques will be used to accomplish this extraction. The objective of data extraction is to measure the similarity of selected user tweets with respect to extracted features. It will enable to determine high percentage of similarity between the user tweets and group characteristics to expose his/her affiliation to this group.