Information Access Using Natural Language

The Problem: With recent advances in computer and Internet technology,peoplehaveaccess to more information than ever before. As the amount of information grows, so does the problem of finding what one is looking for. We believe that the most natural form of communication and information access for humans is natural language. We propose to address the growing information access problem with a uniform natural language interface. Motivation: In December 1993, START (SynTactic Analysis using Reversible Transformations) became the first natural language system available for question answering on the World Wide Web. Since then START has been involved in dialogs with users all over the world, answering millions of questions. As we added more and more information to START’s knowledge base, we discovered the advantages of “virtual collaboration. ” We realized that the existence of the Web, with its huge resources, allows us to put to use the fruits of labor of a large group of people without explicitly collaborating with them. Whenever we findaninteresting knowledge source we can analyze its structure and interface it with START. Then, in response to a question, START can dispatch the user to a weather forecast page, a map collection, a World Factbook database, a personal homepage, etc., all accessed through our uniform natural language interface. It is this “virtual collaboration, ” which became possible only a few years ago, that inspired our project. Previous Work: The project is based on the START natural language system developed by Boris Katz [4, 5]. The