On Modeling Trust in Social Media using Link Polarity1

1 Partial support was provided by NSF awards ITR-IIS-0326460 and ITR-IDM-0219649 and funding from I.B.M. Contact the authors as {akale1,kolari1,aks1,finin,joshi}@cs.umbc.edu ABSTRACT There is a growing interest in exploring the role of social networks to understand how communities and individuals spread influence. In a densely connected online world, social media and networks have a great potential in influencing our thoughts and actions. We describe techniques to model trust in social media and present experimental results on finding “like minded” blogs based on blog-to-blog link sentiment for a particular domain. Using simple sentiment detection techniques, we identify the polarity (positive, negative or neutral) of the text surrounding links that point from one blog post to another. We use a trust propagation model to spread this sentiment from a subset of connected blogs to other blogs and deduce like-minded blogs in the blog graph. We then extend the same technique to label main stream news sources as leftor right-leaning based on the links between blogs and news sources. Our results confirm that the simple heuristics to analyze the text surrounding links and our trust propagation model are highly applicable for domains having weak link structure. These techniques demonstrate the potential of using polar links for more generic problems such as detecting trustworthy nodes in web graphs.