As a guest user you are not logged in or recognized by your IP address. You have
access to the Front Matter, Abstracts, Author Index, Subject Index and the full
text of Open Access publications.
This work aims at creating a user recommender system that recommends relevant people to follow for twitter users. We propose to use a novel topic modeling method Biterm Topic Model (BTM) to profile users into vectors of bag of words. We then propose an algorithm that uses both social network relationship information and the user-generated content modeled through BTM to recommend twitter followees. A preliminary evaluation is carried out on the implementation of this technique that shows BTM performs well in making valid recommendations to twitter users. We also found that considering both user generated content and social relationships for recommending followees helped improve the results.