| Summary: | 碩士 === 國立中央大學 === 資訊管理研究所 === 84 === Chinese input has always been a major problem of Chinese
computing, among many input methods the Zuyin (Phonetic)
input has the advantage of easy-to-learn and easy-to-use,
make it the one worth to improve. And one of our goal is to
develop a workable strategic and method to improve it.
In this paper we try to analysis the methods currently been
used, and try to find out their advantage or shortcoming,to
develop a way to integrate them into our system. We use the
preprocessing, analysis and translation, correction and
learning, three stage to work out a solution. In each stage,
proper analysis has been done to different approach, and a
suggestion has been made to integrate them. The experiment
shows that the final system can perform better in both the
correct rate and processing time.
Besides the processing, the user interface part is also very
important. In this paper, we also analysis different develop
background of Chinese system, and its relation to the input
methods. Also different processing model and different
communication style of input method have been explained,
so does the possible input assistant.
|
|---|
