Automatic Question Generation system

Abstract

The process of automating the question generation consists of many tasks. Selecting the target content (what to ask), question type (who, why, how) and actual question generation are the major issue of Automatic Question Generation. Certain definitions retrieved is available in Wikipedia either directly or is the outcome of executing set of sub queries for each key phrase categories The problem in the existing system is that some of the definition sentences which are taken out from Wikipedia were implicit. The proposed system overcomes the problems by using Supervised Learning Approach, Naïve Bayes method. It also extends its work to use Summarization, Noun Filtering and Question Generation in the aim of generating semantically correct questions.