Correct Real-word Errors through Revising Issues in Contextual Semantic Information

摘要：

Real-word misusage widely exists in English text writing, especially for the essay of those secondlanguage learners. Conventional approaches extract different types of semantic feature from the context to detect and correct the real-word errors, regardless of concerning if these features are validation. This paper, on the other hand, in view of the neglected issue and put forward an innovative method RSA (Real-word Semantic Analysis) on the basis of revising improper context semantic features. In our method, the diversity of word forms in lexicon, word boundary errors, morphological and phonetic confusion are taken in account and a certain amount of candidates are generated to serve as features for real-word disambiguation. The experiment manifests that RSA is capable of providing better performance than Microsoft Word 2007 on real-word disambiguation when test sets involve incorrect contexts.

关键词： context semantic analysis real-word error word boundary confusion phonetic confusion morphological confusion

作者: Guiming Huang Yanzhou Huang

作者单位: Guilin University of Electronic Technology Guilin, China

会议类型: 国际会议

会议名称: 2011 International Conference on Information and Industrial Electronics(2011年信息与工业电子国际会议 ICIIE 2011)

会议地点: 成都

会议语种:英文

页码: 26-30

在线出版日期: 2011-01-14（万方平台首次上网日期，不代表论文的发表时间）