Estimating the Quality of Postings in the Real-time Web

Micro-blogging services, such as Twitter, Jaiku and Plurk, are get- ting increasingly more popular as users are constantly posting in- teresting findings that are made available to other users wit hin sec- onds. In this paper we study the problem of estimating the quality (or "interestingness") of postings in the real-time Web. We identify several important factors that are indicative of the qualit y of post- ings, and present metrics that capture these factors. To showcase the promise of our approach, we present early experimental results on a dataset crawled from Twitter.