The Question Answer System Using Deep Learning

  • Vaishali Fulmali, Prof. K. P. Moholkar

Abstract

Question-answer systems are intelligent systems that can be used to give answers to the user’s questions. A question answering system can be designed that can automatically answer a question, in a similar way as a human can find answers to questions. Community question answering (CQA) services were becoming popular over the past few years. It permits the members of the community to post a question as well as answer the questions. CQA texts are noisy, therefore, the application of the traditional text mining methods does not give good results. In the proposed system, a deep learning approach is used for automatic question- answering. The questions and answers are embedded and the deep neural network is trained to find the similarity between questions. The most relevant answer to the user’s question is provided based on the highest matching score of the user’s question with the questions available in the dataset. The purpose of the proposed system is to design an automated question answering system. The proposed system uses hierarchical clustering to organize the dataset and Bi-LSTM algorithm to find the relevant question.

Published
2020-08-01
Section
Articles