Phonetic Errors Checking of Myanmar Language

Abstract

Spell checker is an essential component of word processing, Machine Translation system and Office Automation system. Myanmar Pronunciation and orthography has differences because spelling is often not an accurate reflection of pronunciation. Pronunciation of a word may lead to misspelling. Beside, in Myanmar compound words, the next word depends on the previous meaning of the word. In this paper, we propose a phonetic error checking algorithm for Myanmar compound words. Firstly, we segment the sentence with longest matching approach. Secondly, we check phonetic error with the proposed algorithm by applying Myanmar corpus. Finally, the system provide suggestion list which contains the correct compound word by computing string cosine similarity. Experimental results showed that proposed approach achieves 88% precision for testing input.