Methods of Automated Question Answering

Abstract

It is human nature to ask questions. Question answering is a very important part of learning. We ask questions to learn more about something and answers help us to gain a better perspective.

The internet is taking over everything in the world. Millions of queries are asked on the internet everyday, most of which are repeated. Hence, it is inefficient to answer the same question repeatedly. We can use the answers of previously asked questions to answer new questions.This can be done by using various Natural Language Processing (NLP) techniques like Bag of Words (BOW), word-to-vec, Glove, BERT, etc.

 In this paper, we try to provide a quantitative analysis between these techniques. We also try to find out what is the best technique to make automated question answering mechanisms.