中文混淆字集应用於别字侦错模板自动产生ChineseConfusionWord.PDFVIP

中文混淆字集應用於別字偵錯模板自動產生 Chinese Confusion Word Set for Automatic Generation of Spelling Error Detecting Template 陳勇志 Yong-Zhi Chen, 吳世弘 Shih-Hung Wu 朝陽科技大學資訊工程系 Department of Computer Science and Information Engineering Chaoyang University of Technology {9727602, shwu}@.tw 盧家慶 Chia-Ching Lu, 谷圳 Tsun Ku 資訊工業策進會 Institute for information industry {gaty, cujing}@.tw 摘要 本研究透過常用字來產生混淆字集，自動產生能夠幫助錯別字偵測的模板，發展華 語文錯別字偵測技術。本系統利用辭典為基礎，使用辭典中的詞彙做為正面用詞，透過 混淆字集自動產生含別字的反面模板，能夠偵測的別字包含同音字、同部首字，並且透 過斷詞軟體輔助擷取更正確的反面模板，用以協助華文教師進行大量華文作文的錯別字 批改甚至輔助學生進行寫作，最後達到提昇寫作能力之成效。 關鍵詞︰模板產生、模板探勘、正反面用語知識庫 Abstract In this research, we proposed a system that can use automatically generated templates for detecting Chinese spelling error. At first, we use frequently used Chinese characters to produce the Chinese confusion set. Based on a dictionary, our system automatically generated negative vocabulary template with the help of Chinese confusion set. Error types include pronunciation-related errors and radical-related errors. And our system uses word segment to capture more accurately the negative template. We hope that such a system can help the teachers on the checking of students’ essays, and also can help students learn to write effectively. Consequently, the students would improve their writing skill. Keywords: Template generation, Template mining, Pragmatics Knowledge Base. 359 一、緒論 自民國95 年起，教育部在國中基本學力測驗中加辦「寫作測驗」隨後列入升學計 分，計分標準依據立意取材、結構組織、遣詞造句、錯別字給予6 個等第的級分，