Modeling blogger influence in a community

Abstract

Blogging has become a popular and convenient way to communicate, publish information, share preferences, voice opinions, provide suggestions, report news, and form virtual communities in the Blogosphere. The blogosphere obeys a power law distribution with very few blogs being extremely influential and a huge number of blogs being largely unknown. Regardless of a (multi-author) blog being influential or not, there are influential bloggers. However, the sheer number of such blogs makes it extremely challenging to study each one of them. One way to analyze these blogs is to find influential bloggers and consider them as the community representatives. Influential bloggers can impact fellow bloggers in various ways. In this paper, we study the problem of identifying influential bloggers. We define influential bloggers, investigate their characteristics, discuss the challenges with identification, develop a model to quantify their influence, and pave the way for further research leading to more sophisticated models that enable categorization of various types of influential bloggers. To highlight these issues, we conduct experiments using data from blogs, evaluate multiple facets of the problem, and present a unique and objective evaluation strategy given the subjectivity in defining the influence, in addition to various other analytical capabilities. We conclude with interesting findings and future work.