A New Approach to Blog Post Summarization Using Fast Features
Abstract: Blog post summarization using fast features facilitates users' quick browsing through blog search results. Much existing research on blogs ignores blog tags and text structure. In this paper, we re-formalize the blog post summarization problem as a sentence extraction and sentence ranking problem. Three fast features, important sentences, blog tags and blog comments, are proposed to calculate salience scores of representative words. Then we propose an average-summation-based sentence selection method called ASS to select sentences based on the salience scores of content words in sentences. As been used to evaluate on human-labeled sentences, ASS showed anticipated results.
Loading