We develop a system for mining events and unseen name translations from online daily Web news. This system first automatically discovers bilingual events by analyzing the content of the news stories. The discovered event can be treated as comparable bilingual news and can be used for generating name candidates. A name matching algorithm is developed to discover new unseen name translations based on phonetic and context clues. The experimental results show that our system is effective for mining new knowledge and information from online Web news.
Published in:
Digital Libraries, 2004. Proceedings of the 2004 Joint ACM/IEEE Conference on
Date of Conference: 7-11 June 2004