Identifying breakpoints in public opinion

Abstract

While polls are traditionally used for observing public opinion, they provide a point snapshot, not a continuum. We consider the problem of identifying breakpoints in public opinion, and propose using micro-blogging sites to capture trends in public opinion. We develop methods to detect changes in public opinion, and find events that cause these changes.Our experiments show that the proposed methods are able to determine changes in public opinion and extract the major news about the events effectively. We also deploy an application where users can view the important news stories for a continuing event and find the related articles on web.