Parts-of-Speech Tagger for Gujarati Language using Long-short-Term-Memory

Abstract

Parts-of-Speech (POS) tagging is a crucial step to process the natural languages. It is a state-of-art method of providing the lexicon category such as noun, verb, adjective, etc. to each word that best suits the context of the sentence in which it is used. Being a part of pre-processing makes this task an important step in linguistics and semantics. Gujarati is an Indian language widely spoken in Asia and across the world. Part-of-Speech tagging can be used in word sense disambiguation, Information retrieval, machine translation and parsing. In this paper, we proposed Long-short-Term-Memory (LSTM) based Part-of-Speech tagger for Gujarati language. With our proposed approach, this paper envisions achieving accuracy of 95.34% and 96% precision with the help of this novel & efficient gradient based method.