The goal of this SBIR Phase I project is to develop software that will allow health professionals to search for health science information using natural language questions submitted over the Internet, as well as to manage materials and processes ate their local laboratories and clinics. This project will build on the success of the Data/Knowledge Bases and Keynet Projects at Northeastern University. The proposed information access mechanism will use the search engine technology developed in the Keynet Project. This search engine is inexpensive, yet has high performance and rapid responsiveness; moreover, it is designed to be used with structure information rather than ordinary flat text. In particular, it can support information retrieval of nontextual objects such as images. The proposed search engine will use the Unified Medical Language System as its basis. PROPOSED COMMERCIAL APPLICATION: The health sciences constitute a large market for software, yet there are very few examples of knowledge-based software available. Accordingly there is a large potential market, even if one only considers small research labs in academic institutions and in small companies NATIONAL LIBRARY OF MEDICINE
