Interaction-Based Behavioral Analysis in Twitter Social Network

Literature studies usually use data sets consisting of data collected from many different metrics and user counts collected over different time periods. The data set used in this article was formed using completely up-to-date data obtained as a result of metrics measured in terms of scope and efficiency, sufficient and effective user counts, and filtering processes. To classify users correctly and make the classification performance high—in addition to parameters used in the literature such as tweets, account age, follower rank, average retweets and average likes—other parameters such as diameter, density, reciprocity, centralization and modularity were used. These metrics are the parameters that focus on a different area to reveal many aspects in which social network users interact. The data used to create the data set was collected from Twitter. The metric data forming the data set was extracted using Twitter Rest API V1.1 supporting search/tweet endpoints by means of the SocialBlade and Netlytic platforms.