16 Multi-Clustering Users in Twitter Dataset

There are significant numbers of users in different social networks which are increasing every day. People join different social networks for different purposes. The diverse possibilities which these social networks provide, have resulted in majority of users chose these social networks as a virtual place to talk, discuss and share different things in diverse fields and interests. This traffic increasing in using these websites, have caused that many researchers and software engineers, focus on content and structure of these networks. This work aims to cluster users in Twitter dataset to find the more discussed keywords among connected users. In order to have accurate clusters, in this work users are clustered based on common keyword and the connectivity among them in an identifiable interval and finally clusters are ranked based on the number of users and number of common keywords used among them. Taking both keyword similarity and network structure into account for clustering users leads to a more meaningful clusters.