Deep Learning Techniques and Models for improving Machine Reading Comprehension System

  • Archana Nalavade, Anita Bai, Megha Bhushan

Abstract

Machine reading comprehension (MRC) is to answer queries based on natural language documents. It is a new challenging direction in Artificial Intelligence. MRC can be viewed as a task of reading a piece of text written in natural language, understanding it, and answering the associated question about the corresponding context. The evolution in deep learning techniques in addition to the availability around the large datasets has made the MRC very successful in the early years. MRC is a rising field of research due to its potential appliance in a variety of enterprise uses and applications, as well as the availability of MRC benchmarking datasets (MSMARCO, SQuAD, NewsQA, etc.). In this paper we have included recent advances in MRC, highlighting aspects such as MRC datasets, fundamental tasks in machine reading comprehension, wide-ranging structural design of neural MRC with the most important modules and common methodologies along with techniques to carry out the tasks in MRC.

Published
2020-10-03
How to Cite
Archana Nalavade, Anita Bai, Megha Bhushan. (2020). Deep Learning Techniques and Models for improving Machine Reading Comprehension System. International Journal of Advanced Science and Technology, 29(04), 9692 -. Retrieved from http://sersc.org/journals/index.php/IJAST/article/view/32996