Question answering using deep learning in low resource Indian language Marathi

TL;DR

This paper explores the use of transformer models for Marathi question answering, fine-tuning multilingual and monolingual models to improve accuracy in a low-resource language setting.

Contribution

It evaluates various transformer models for Marathi QA and demonstrates the effectiveness of MuRIL in achieving high accuracy on Marathi reading comprehension tasks.

Findings

MuRIL achieved an EM score of 0.64 and F1 score of 0.74.

Transformer models can effectively address low-resource language QA.

Fine-tuning multilingual models improves Marathi QA performance.

Abstract

Precise answers are extracted from a text for a given input question in a question answering system. Marathi question answering system is created in recent studies by using ontology, rule base and machine learning based approaches. Recently transformer models and transfer learning approaches are used to solve question answering challenges. In this paper we investigate different transformer models for creating a reading comprehension-based Marathi question answering system. We have experimented on different pretrained Marathi language multilingual and monolingual models like Multilingual Representations for Indian Languages (MuRIL), MahaBERT, Indic Bidirectional Encoder Representations from Transformers (IndicBERT) and fine-tuned it on a Marathi reading comprehension-based data set. We got the best accuracy in a MuRIL multilingual model with an EM score of 0.64 and F1 score of 0.74 by fine tuning the model on the Marathi dataset.