Auto-FAQ-gen: Automatic frequently asked questions generation

Abstract

Using Frequently Asked Questions (FAQs) is a popular way of documenting the list of common questions on particular topics or specific contexts. Most FAQ pages on the Internet are static and can quickly become outdated. We propose to extend the existing work on question answering systems to generating FAQ lists. In conventional Question Answering (QA) systems, users are only allowed to express their queries in a natural language format. A new methodology is proposed to construct the questions by combining of question extraction and question generation methods. The proposed system accepts, extracts, or generates user questions in order to create, maintain, and improve the FAQs quality. In addition to present the basis of QA system, complimentary units will be added to conduct the FAQ list. The research proposed here will contribute to the field of Natural Language Processing, Text Mining, QA, particularly to provide high quality automatic FAQ generation and retrieval.