Abstract:
In this study, author proposed a concept-based recommender of followees in
Twitter. The task is performed by mining user interests represented a set of
salient concepts in users tweets.
This to support and facilitate the knowledge discovery process from user tweets.
The process starts with fetching process of users
last 100 tweets representing his/her latest interest. The system then processes
them further to discover those salient concepts of the user. Following that,
it identifies those users with similar interest. Similarities between Twitter
users are calculated using clustering of users with similar interests. Initial
experiments show that the proposed technique is able to perform the task effectively.