Conference Proceedings
Knowledge discovery in sportsfinder: An agent to extract sports results from the web
H Lu, L Sterling, A Wyatt
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) | Published : 1999
Abstract
There is a wealth of information to be mined from the World Wide Web. Unfortunately, standard natural language processing (NLP) extraction techniques perform poorly on the choppy, semi-structured information fragments, such as sports results, which are popular to be published on the Web pages nowadays. In this paper, we present an information agent: SportsFinder, an agent to extract sports scores from the World Wide Web, as well as the knowledge discovering method to learn new express patterns to improve the agent’s performance.