Discovering Knowledge from Conference Web Pages

This paper proposes an approach to discover knowledge from conference web pages and organize such knowledge as ontologies. Firstly, conference web pages are segmented into blocks by analyzing visual features and the DOM structure. Then we use Bayes Network to classify the blocks into predefined categories. Finally, ontologies are generated from the classified text blocks. The experimental results on real-world datasets show that the proposed method is effective and efficient for discovering ontology concepts and hierarchies from conference web pages.