As the growth of exchange activities between four regions of cross strait, the problem to correctly convert between Traditional Chinese (TC) and Simplified Chinese (SC) is getting important and attention from many people, especially in business organizations and translation companies. Different from the approaches of many conventional code conversion systems, which rely on various levels of human constructed knowledge (from character set to semantic level) to facilitate the translation purpose, this paper proposes a Chinese conversion model based on Maximum Entropy (ME), a Machine Learning (ML) technique. This approach uses tagged corpus as the only information source for creating the conversion model. The constructed model is evaluated with selected ambiguous characters to investigate the recall rate as well as the conversion accuracy. The experiment results show that the proposed model is comparable to the state of the art conversion system.