Improving Relevance of Keyword Extraction from the Web Utilizing Visual Style Information

摘要

Information growth is faster than ever before. We need to provide advanced services facilitating information "consumption" (e.g., recommendation, personalized navigation). At least a lightweight semantics is necessary for such services. Nowadays keyword paradigm is widely used and seems to achieve satisfactory results in fields such as social bookmarking or ontology learning. In this paper we explore impact of web site visual style on relevant keywords extraction. We propose a method for relevant keywords extraction from web pages combining traditional automatic term recognition algorithms with web site's visual style processing. We particularly focus on cascade style sheets. The evaluation conducted on 200 "wild" Web documents from 12 different web sites showed that our method increases the relevance of extracted keywords.