Blog classification using K-means

With the recent exponential growth of blogs, a vast amount of important data has appeared on blogs. However, dynamic, autonomous, and personal features of such blogs make blog pages be quite different from those on general web pages in many aspects. As a result, this also causes many problems which cannot be handled properly by general search engines. One of the problems which we focused in this study is that blog pages are inherently poorly-organized and very much duplicated. This means the blog search engines cannot but provide the poorly-organized and duplicated results. To solve this problem, we propose a blog classification method using K-means and present a blog search result reorganization approach based on this method. In this study, firstly, we review the current status and their performances of blogs and blog search engines. Secondly, we adopt the K-means algorithm as a base algorithm and devise a blog title classification method to reorganize the blog titles resulted by a search engine. Finally, by implementing a prototype system of our algorithm, we evaluate our algorithm's effectiveness, and present a conclusion and the directions for future work. We expect this algorithm can improve the current blog search engines' usability.