Partially Constructed Knowledge for Semantic Query

• Main Authors: Ching-Lung Chen, 陳慶龍
• Other Authors: Tzone-I Wang
• Format: Others
• Language: zh-TW
• Published: 2007
• Online Access: http://ndltd.ncl.edu.tw/handle/11685266672222851466

碩士 === 國立成功大學 === 工程科學系碩博士班 === 95 === With the development of World Wide Web (WWW), huge amounts of data are exchanged on the Internet. How to reuse knowledge correctly, effectively and conveniently is an important issue. Nowadays, the most popular tools for seeking the knowledge in network are Google or Yahoo keyword-based search engine. After obtaining the documents of webpage which gained from search engine, users have to read and organize it themselves. If the information you obtained was extremely huge, it may take too much time to organize it easily. The question answering system (QA system) provides the most appropriate answer for user by giving a question. It not only save the time but also organize and reuse knowledge for users. However, it is difficult for QA system to understand the language semantic and to extract knowledge from question and source documents. In this thesis, we design the dynamically automatic ontology constructor for semantic search. For answering user question, we construct partial sports ontology by expanding the semantic of question. After the ontology is built, the system can infer the answer without asking users to read the whole document sets. NBA and CPBL sport news pages are used as source documents in our research, while the approach can be used in other domain.