The authors propose a personalized blog reader (PBR) system that helps bloggers find interesting posts. Their proposal includes an online incremental clustering algorithm that uses blog link and content information, and, for personalized blog ranking, a method to learn bloggers’ reading habits.

Li et al. measure various aspects of their system. For example, they assess the consistency of the results of various possible clustering approaches with a ground truth. For this purpose, they use the Rand index, although a well-known and more reliable version of this statistic exists that eliminates the chance factor [1]. The authors also evaluate the effects of some parameters on the clustering runtime efficiency, and present the results of a user study on the subjective evaluation of their system.

This fast-paced paper introduces many things. It includes some unnecessary details, such as the formula for the Rand index. Also, some of the sentences are awkward and long.

The paper is interesting. Blogging is a popular Web activity, and studies such as this one are needed.