Conversational Question Answering System using RASA Framework

Abstract

Conversational question answering systems areaway ofhuman-computer interaction which integrates the automation into everyday life. Chatbots gained popularity because of their ease and simplicity, which helps the user to resolve their queries. In this work, we developed a conversational question answering system with a RASA framework to handle students’ queries regarding entrance exams and handle general informative questions. The system Classifies the intent and entities of the user input and calculates the confidence with respect to the training data. If confidence falls below a threshold, ask the user to reframe the query. This chatbot currently supports nine languages for user questions and is an excellent means to resolve queries of enthusiastic students.