An efficient & learning approach of POS tagging using rule-based for devanagari script

Abstract

In this paper, we present a novel approach for Devanagari Script which can help in various applications like:- Named Entity Resolution, Sentiment Analysis, and Question Answer. This framework works on the following approaches: Learning Task, Performance Efficiency Improvement Task. This system will automatically learn its rules and the main focus is on the tagger which automatically identifies, & resolves its shortcomings which will incrementally increased the performance. The basic problems with POS tagging for Indian Languages are very difficult due to some constraints in there. But with some rules and overcoming of limitations restrictions this can be solved. User provides the sentence which goes through a series of stages. In each stage, system also learns the new rules and provides the better results.