A Simple Rule based Myanmar POS Tagger

Abstract

This paper presents a simple rule-based POS tagger to tag the correct syntactic categories of the Myanmar words by applying lexicon based word segmentation and heuristic rule based tagging method. Firstly, input sentence is tokenized into words by using syllable breaking and syllable merging with longest matching approach. Secondly, this system defines the detailed tag sets for POS tagging process by using nine different POS in Myanmar grammar. Finally, the proposed system solves the POS ambiguities of one word by applying word disambiguation rules which are generated from morphological features of Myanmar grammar. So, the proposed system can provide many benefits to Myanmar-English translation system and other NLP tasks.