Abstract—Users provide a lot of undiscovered information when they are working with electronic products, such as computer. Many natural annotations are left when users type Chinese materials with input method. We call them natural typing annotations. These annotations have intimate connection with Chinese word segmentation. In this paper we collect corpus with natural inputting annotations and analyze it in various respects. From the corpus, user’s typing patterns are extracted and classification model is built to identify different patterns. Experiments show that natural inputting annotations have promising potential in overcoming the drawbacks of existing word segmentation approaches.
Index Terms—Natural inputting annotations, Chinese word segmentation, user’s typing patterns, classification model, voting mechanism.
The authors are with the Beijing Institute of Technology, Beijing 100081, China (e-mail: dakuiz@163.com; maoyu1998@qq.com).
[PDF]
Cite: Dakui Zhang and Yu Mao, "Typing Behaviour Gives Chinese Word Segmentation New Perspective," International Journal of Engineering and Technology vol. 9, no. 6, pp. 498-502, 2017.
