Paper Title
MULTILINGUAL QUESTION AND ANSWERING FRAMEWORK
Abstract
Abstract - A Multiple language question answering framework is an important part of a natural language chatbot system.But existing approaches show performance that is much inferior to that of monolingual systems. In this paper, we have proposed a novel approach called a Multilingual Question-answering framework for the effective handling of user queries in a chat interface primarily targetedat solving HR queries. The user can query in a selection of Indian languages. Our approach includes a pretrained M-Bert model which is trained on the XSQUAD dataset. The response will be either a sentence (span of text) or a response related to the context. The model gives the relevant answer with good accuracy when compared to other translation models.
Keywords - Multilingual Natural Language Chat Interface, Question- Answering, Multilingual Question Answering Framework, Cross-Lingual Question Answering.