未登录词识别是汉语分词处理中的一个难点。在大规模中文文本的自动分词处理中,未登录词是造成分词错误的一个重要原因。为了解决自动分词的这一“瓶颈”问题,对未登录词中的新词语的结构进行分析,总结新词语的构词类序,发掘新词语的构词规律,以期为未登录词的识别和标注提供一套规则。 Unrecognized word recognition is a difficult point in Chinese word segmentation. In large-scale Chinese text automatic word segmentation, unregistered words is an important reason for word segmentation errors. In order to solve the problem of “bottleneck” in automatic word segmentation, this paper analyzes the structure of new words in unregistered words, summarizes the word-formation order of new words, and excavates the word formation rules of new words, in order to identify unregistered words And labeling provide a set of rules.