Investigation of mechanisms to improve question answering

This document is describes research in Internet Frequently Asked Question (FAQ) answering and Natural Language Processing (NLP). The research seeks to enhance the accuracy of FAQ file answer searches of an existing FAQ answering system. Secondly, this research seeks to increase breadth through implementation of a WEB search and classification mechanism. Accuracy improvements are intended to provide an increased ability to correctly match a user question with a corpus of previously asked questions residing in Internet FAQ files. Breadth increases will be accomplished through an automated approach to scanning and categorization of Internet FAQ files.