Chinese Word Segmentation

Chinese word segmentation has been a very important research topic not only because it is usually the very first step for Chinese text processing, but also because its high accuracy is a prerequisite for a high performance Chinese text processing such as Chinese input, speech recognition, machine translation and language understanding, etc. This paper gives a review on the development of Chinese word segmentation techniques that have been applied to various applications on Chinese text processing. As the methodology varies in a very wide range according to its applications, in this paper it is viewed in terms of the knowledge resources on which segmentation methods based. We summarize the methods into two categories, that is, lexical knowledge based and linguistic knowledge based methods.