Post Ranking in a Blogosphere: Algorithms and Evaluation

Abstract

Existing algorithms of post ranking in a blogosphere do not consider the information on categories and timestamps of posts, which is valuable to improve the accuracy of post rankings. The category, created by a blogger, implies her/his interesting topics, and the timestamps, which indicates the creation times of a post and a blog action, are closely related to the quality of a post. In this paper, we propose three strategies that exploit the information related to the categories and timestamps, and address how to apply it to four existing post ranking algorithms, i.e., PostRank, BAITS, BloggerAVG, and PBFS. We evaluate the accuracy of these algorithms by conducting extensive experiments using real-world blog data. The results show that our strategies help each algorithm improve the accuracy. (1)