Hybrid Approach for Construction of Summaries and Clusters of Blog Data for Improved Blog Search Results

The data are noisy in blogs, because blog entries are unstructured and might cover a wide variety of topics. However, because of the number of blogs exist, manually viewing and examining them is a difficult and time-consuming task. Intuitively, you could apply existing text and Web mining techniques to blog analysis and mining. But, because of the existence of various challenges, we can’t directly apply these techniques. The bloggers update their information content on the blogs much more frequently than Web masters update traditional Web pages, often daily or even hourly. Above all, bloggers cover very diverse topics, so maybe only one paragraph in a particular entry could relate to someone’s topic of interest. In this paper we propose an architecture which takes a query from the user, process through blog parser and extract content from the blog page. Then we identify the sentences which should be taken for further processing using a blog analyzer and finally summarizing the content based on the analysis results. The process is repeated for all the blogs and results in the summarized output of clustered blogs.