This Small Business Innovation Research Phase II project from Teragram Corporation embodies the research and development in information retrieval and natural language processing that will locate within free text the precise answers to direct English questions, without relying on the traditional and unnatural search methods of keywords and Boolean search strings. The project will yield a new generation of searching and indexing technologies, one that will empower users to target existing answers with intuitive full sentence queries. In Phase II, Teragram will build a Question and Answer search engine whose performance will be inherently better than the generalized search engines in use today. Strategies to capture English question-and-answer pairings will be automatic, precise, flexible, and scaleable to enable the mastery of the very large information resources now common on the internetl and many intranets as well. The engine's indexing and matching processes will be uniquely sensitive to Q&A content so as to ensure the richest lookups possible. The engine will comb the entire World Wide Web or just one site of Frequently Asked Questions (FAQS) with equal ease. The effect will be to make the expert behind any good FAQs list, or similarly structured knowledge, come marvelously to life for the user's immediate benefits. Teragram Corporation developed natural language search tools, which have already yielded a large customer base. These powerful functions have been successfully deployed by a major Internet search engine. The research and development of Phase II for a Question-Answer search and indexing engine important for fact retrieval will have enormous commercial implications for Internet search engines, corporate Intranets, technical support systems, and other large scale documentation environments.
