Learning the Semantic Meaning of a Concept from the WebTweet
Speaker: Yang Yu
Start: Thursday, August 03, 2006, 01:00PM
End: Thursday, August 03, 2006, 03:00PM
Location: 325b ITE, UMBC
Abstract: Many researchers have applied text classification techniques to the ontology mapping problem. The mapping results in these researches heavily depend on the availability of highly relevant text exemplars associated with individual concepts. However, manual preparation of exemplars is costly. In this work, we propose to automatically collect text exemplars by downloading and processing web pages listed in the search results obtained by querying a search engine. Search queries are formed for each concept according to the semantic information given in the ontology. We have implemented a prototype system and conducted a series of experiments. Given two ontologies, the process from forming search queries to calculating conditional probability of two concepts is fully automated. We assessed the effectiveness of our approach by comparing the obtained conditional probabilities in these experiments with human expectations. Our main contribution is that we explored the possibilities of utilizing web information for text classification based on-tology mapping and made several valuable discoveries on its usefulness for future research.
Web Site: www.csee.umbc.edu/~yangu1
Host: Yun Peng