Portal de Periódicos da CAPES

Blog Ranking Based on Bloggers’ Knowledge Level for Providing Credible Information

2009; Springer Science+Business Media; Linguagem: Inglês

10.1007/978-3-642-04409-0_26

1611-3349

Shinsuke Nakajima, Jianwei Zhang, Yoichi Inagaki, Tomoaki Kusano, Reyn Nakamoto,

Topic Modeling

With the huge increase of recently popular user-generated content on the Web, searching for credible information has become progressively difficult. In this paper, we focus on blogs, one kind of user-generated content, and propose a credibility-focused blog ranking method based on bloggers' knowledge level. This method calculates knowledge scores for bloggers and ranks blog entries based on bloggers' knowledge level. Bloggers' knowledge level is evaluated based on their usage of domain-specific words in their past blog entries. A blogger is given multiple scores with respect to various topic areas. In our method, blog entries written by knowledgeable bloggers have higher rankings than those written by common bloggers. Additionally, our system can present multiple ranking lists of blog entries from the perspectives of different bloggers' groups. This allows users to estimate the trustworthiness of blog contents from multiple aspects. We built a prototype of the proposed system, and our experimental evaluation showed that our method could effectively rank bloggers and blog entries.