Punjabi Chunker using Bootstrapping Approach

Abstract

This paper represents Punjabi Chunker using bootstrapping approach. Bootstrapping is an approach which does not need any external input that’s why it is also known as self-starting process. It is semi-supervised technique in which collection of both labeled and unlabeled data is taken. It helps to make use of unlabeled data by training a small amount of labeled data. Semi supervised learning is fall in the middle of supervised learning and unsupervised learning Chunking is the process of breaking long strings of information into units or chunks. Chunking is different from parsing. POS, Named entity Recognition and sentence breaking are the main applications of NLP in which chunking are used. This research work is different from greedy algorithm because in this approach both labeled (trained) and unlabeled data set is used to built text Chunker for Punjabi language.

Keywords: Natural language Processing (NLP), Part of Speech Tagger (POS), Punjabi Chunker.