Abstract: The concept of relative word frequency (RWF) is proposed. A context calculation model is set up, which makes use of contextual information to resolute covering ambiguity in Chinese word segmentation. This paper selects 5 frequently used covering ambiguous words as examples, and the results show that the average accuracy is over 95%.

Key words: Chinese word segmentation, Covering ambiguity, Relative word frequency, Context calculation model

摘要： 提出了相对词频的概念，据此建立了语境计算模型，利用歧义字段前后语境信息对组合型分词歧义进行消解。对高频出现的5个组合型分词歧义进行实验，平均准确率达到95%以上，证明该方法对于消解组合型分词歧义具有良好效果。

关键词: 中文自动分词, 组合型歧义, 相对词频, 语境计算模型

CLC Number: 

• TP311