Finding influential users and popular contents on twitter

Abstract

On Twitter, People do not only find new friends by following others, but also propagation the information by retweeting. So, we can not measure the users’ influence only by following relationships easily, also, it is not reasonable to measure tweets’ popularity by the number of retweets. In this paper, a novel random walk model was proposed to measure the users’ influence and tweets’ popularity. In our model, the influence of users was measured not only by random walk of the following network, but also by the popularity of tweets. In fact, if a user often tweets popular contents firstly, we think this user is important and the influence of the user is higher. Moreover, if a content is retweeted by many high influencers, we think this content is important and popular. Experiments were conducted on a real dataset from Twitter containing about 0.26 million users and 10 million tweets, and results show that our method is consistently better than PageRank method with the network of following and the method of retweetNum which measures the popularity of contents according to the number of retweets.