A systematic exposition of Punjabi Named Entity Recognition using different Machine Learning models

Named Entity Recognition plays a vital role in diverse Natural Language Processing applications like Information Extraction, Information Retrieval, Text Mining, Machine Translation, and Question Answering etc. This paper presents the evaluation of Named Entity Recognition task for a resource poor language like Punjabi. Different challenges posed in the task have also been discussed. A annotated corpus of 2,00,000 words have been developed for recognition task. The system has been evaluated on different machine learning models like Hidden Markov Model, Maximum Entropy and Conditional Random Fields with f-score values of 77.61, 83.65 and 93.21 respectively.