Knowledge discovery in the blogosphere: Approaches and challenges

Abstract

Knowledge discovery in blogs is different from knowledge discovery in areas such as databases or Web documents due to blogs' unique characteristics, which introduce additional mining challenges. Although researchers have investigated several techniques to address different aspects of blog discovery, no comparisons among key knowledge discovery techniques for blogs exist. This article examines three prominent techniques that are frequently applied to discovery in blogs - clustering, matrix decomposition, and ranking. The authors compare them in terms of effectiveness in combating present challenges and their ability to accomplish challenging tasks required for effective blog mining. © 2010 IEEE.