頁籤選單縮合
題名 | 漢語社會科學同義詞字面相似度測試--「社會科學檢索詞表」的統計分析=Measurement of Value of Word--Form Similarity Among the Chinese Synonyms in Social Science; Statistics & Analysis of Social Science Thesaurofacet |
---|---|
作者 | 侯漢清; 吳桂英; | 書刊名 | 中國圖書館學會會報 |
卷期 | 63 1999.12[民88.12] |
頁次 | 頁127-133 |
分類號 | 028.85 |
關鍵詞 | 社會科學; 同義詞識別; 字面相似度; 自然語言檢索; Social science; Synonym identification; Word-form similarity; Natural language retrieval; |
語文 | 中文(Chinese) |
中文摘要 | 計算機識別同義詞和準同義詞,是編製後控制詞表、實現目然語言檢索的關鍵技 術,本文以《社會科學檢索詞表》中3725組同義詞為統計分析對象,發現絕大多 數社會科學領域中的同義詞存在著一定的字面相似,近80﹪的同義詞字面相似度 大於(等於)0.50。論文認為,利用字面相似度計算機輔助識別漢語同義詞的方 法是可行的,讓閾值可控制在0.50。 |
英文摘要 | Identification of Chinese synonyms by computer is the main means of using post-controlled vocabulary and natural language in information retrieval. This paper samples and analyzes 3725 Chinese synonyms from the Social Science Thesaurofacet, and finds the 80﹪ of Chinese synonyms in social sciences have word-form similarity and value of word-form similarity of >=0.50. This paper draws a conclusion that it is feasible to identify Chinese synonyms with the aid of computers. Finally it suggests that the threshold value of word-form similarity within 0.5 is the most effective. |
本系統之摘要資訊系依該期刊論文摘要之資訊為主。