An analysis of network structure and post content for blog post recommendation

Abstract

The acceleration of Weblogs has increased the perceived information overload for bloggers attempting to find interested or relevant information. Helping bloggers to efficiently locate relevant and high-quality information is imperative. In this research, we therefore propose four approaches that exploit the post citation network, blog-based social network, and post content to facilitate the automatic construction of an authoritative blog post recommender system. The proposed approaches were tested with blog data collected from Baidu Space, and the experimental results revealed that the proposed approaches outperform the content-only approach and the explicit citation approach.