iTextMM: Intelligent Text Input System for Myanmar Language on Android Smartphone

摘要

In recent years, there are huge developments in mobile phone communication technology to 3G. 3G mobile phone in present form has offered to use internet and interact with the computing system in users own language. However, the efficient input method for Myanmar Language that can be used in 3G mobile phones is still a biggest issue. Moreover, the service of character or word prediction text input system is provided for English and other languages. This paper tried to figure out the development of an innovative Myanmar syllable prediction text input system for Android touch screen mobile phones that leverages structural information of Myanmar characters formation and statistical properties of lexicon resources. In iTextMM, by using position aware rule based matching algorithm and bigram model that achieved a desirable inputting performance compared with currently used prevalent mobile Myanmar input method on Android touch phone (MyanDroid). iTextMM has been released to public via Android Market and is currently in use by hundreds of native Myanmar Android smart phone users. An evaluation results show that the proposed method outperforms the conventional Myanmar text inputting method, approximately 50% in inputting performance.