陳光華2006-07-262018-05-302006-07-262018-05-302002-07-31http://ntur.lib.ntu.edu.tw//handle/246246/20413本計畫第一年已完成基本的控制詞彙指派機制，第 二年完成分類機制的建立。第三年則利用第一年與 第二年的成果，建構一個運用控制詞彙進行檢索系 統。控制詞彙的使用主要為1)取代使用者的查詢 詞彙；2)擴展使用者的查詢詞彙。本計畫將著重於 擴展查詢詞彙，這種作法可以一併考慮使用者的詞 彙。擴展查詢詞彙的方式有：以控制詞彙擴展；以 同義詞典擴展，以控制詞彙加權。除此之外，控制 詞彙之間的關係亦是本研究探究的重點。系統績效 的評比則是以不經任何處理的使用者查詢問句作 為基線模式（Baseline Model），比較前述擴展方式 的績效，以探究控制詞彙的效用。The first year of this project has built an indexing module; .the second year has built a classification module. The report will describe the construction of a retrieval system with the capability of using controlled vocabulary. The application of controlled vocabulary is to replace or to expand the free terms submitted by users. The report focuses on the expansion of user’s free terms, since this method could introduce the capabilities of free terms and controlled vocabulary. Tow kind of query expansions are used in this research. The first uses controlled vocabulary only. The second uses statistical synonym first and then uses controlled vocabulary. In addition, the relations of controlled vocabulary are also considered in the research, e.g., the broad term, narrow term, and related term. A baseline model is used to compare the performance of query expansion.application/pdf301665 bytesapplication/pdfzh-TW國立臺灣大學圖書資訊學系暨研究所reporthttp://ntur.lib.ntu.edu.tw/bitstream/246246/20413/1/902213E002047.pdf