Twitter News: Harnessing Twitter to Build an Article Recommendation System

Abstract:

With more than 140 million active users and 340 million tweets a day (as of March 2012), Twitter presents a great source of recommendation knowledge for articles shared on the platform. In this work, we analyze 836 Twitter users from the technology and entrepreneurship domain with 78,508 links shared by them. We explore and evaluate different (existing and novel) techniques for a recommender system for articles including the following ones: vector-to-vector similarity where the user vector is constructed from the text of the tweets produced, topic-modeling based approach where we learn the topic distribution for each article, as well as the novel hybrid technique which is based on piecewise article vector representation, content-boosted collaborative filtering with pseudo user-ratings as well as the relatedness function which depends on relevance, novelty, connection size and transition smoothness
between articles.