POS Tagging of Assamese Language and Performance Analysis of CRF++ and fnTBL Approaches

Abstract

Assamese is one of the regional languages of India spoken by the people of Assam and other north eastern states of India. Parts Of Speech (POS) tagging is one of the most important research issue as it is the basic need for any Natural Language Processing (NLP). An automated way to provide a Parts Of Speech label to a word on a context is known as Parts Of Speech Tagging. Assamese is one, among the less computationally aware languages of India. This paper presents our works on POS tagging for Assamese sentences, using Conditional Random Field (CRF) and Transformation Based Learning (TBL). We obtain 87.17 and 67.73 percent tagging accuracy for TBL and CRF respectively that are train through a manually tagged corpus.