Abstract: There is a rapid growth on Information Technology through e-Government concept and the usage of mobile phone is also increasing day by day. The majority of SMS writers in Sri Lanka transliterate messages because of language barriers, which create various communication problems and ambiguity of messages. Computing in Sinhala language is an emerging trend in Sri Lanka. This research is an attempt to predict Sinhala sentences in mobile short messages. This is a timely necessity in Sri Lanka.