本文考察了《现代汉语学习词典》等几部辞书的词性标注分歧 ,认为各家出现分歧的原因在于“依据标准不统一 ,语感存在差异”。据此提出以客观的频率统计的结果做标准来解决词性标注问题 ,并阐述了分类依据与划界标准的区分 ,论证了频率统计的可行性与科学性。 This dissertation examines the dissenting part-of-speech of several dictionaries such as “Dictionary of Modern Chinese Learners”, and concludes that the reasons for their disagreements lie in “the inconsistency based on the standard and the difference in linguistic sense”. Based on this, we propose to solve the problem of part-of-speech tagging by using the result of objective frequency statistics. The distinction between classification criteria and demarcation criteria is expounded. The feasibility and scientificity of frequency statistics are demonstrated.