【摘要】 针对已有的词义消歧研究方法的不足,在分析了多种不同结构知识词典的可计算性及其计算复杂度之后,选择北大计算语言所的《现代汉语语法信息词典》和《现代汉语语义词典》,并结合已经标注了词义的人民日报语料作为词义消歧知识源,从中获取汉语词义消歧所需要的统计知识和规则知识,并采用统计与规则相结合的方法构建词义消歧模型,取得了比较满意的词义消其效果。更多还原

【Abstract】 In the light of the disadvantage of word sense disambiguation(WSD) research method,the grammatical knowledge-base of contemporary Chinese and dictionary of temporary Chinese semanitcs are chosen from the Institute of Computational Chinese linguistics of Peking University,after a series of analysis of computability and computational complexity of knowledge dictionary with different structures.Combined the People’s Daily corpus,on which has been tagged word sense,as a word sense disambiguation knowledge source,statistical knowledge and rules knowledge are obtained,which are needed by the Chinese word sense disambiguation.An approach of rule and statistics to construct the model of word sense disambiguation is adopted with satisfactory effect.更多还原