URL Recommendation System in Twitter Using User Activity and User Influence
Seong-Yun Lee, Taewhi Lee, Hyoung-Joo Kim · 2011
Twitter is one of the most attractive microblogging services and has been used as information net-work by people worldwide. But since the number of users is increasing rapidly, it is hard for each individual user to find useful contents from the huge amount of information generated by the users. To solve this problem, URL link recommendation system for Twitter has been proposed, but no previous work has considered the characteristics of Twitter itself. In this paper, we improve the existing system by considering user activity and user influence. We explore factors that improve the quality of recommendations and design methods lo measure user activity and influence. We propose a novel URL recommendation system that combines these factors with the existing system. The experiments show that our system increases user satisfaction.