Korean part-of-speech tagging based on context information

Abstract

This paper describes a part-of-speech tagging system using both rules and statistical information based on context information between an ambiguous word and words around it to resolve lexical ambiguities. The rules and statistical information are composed of POS tag and/or morpheme for an ambiguous word, and POS tags and/or morphemes for words around it. The system apply rules first, and then applies statistical information when the ambiguities have not been resolved. Experimental results shows that our POS tagging system has high-accuracy and broad coverage.