Robin Burke, Kristian Hammond and Julia Kozlovsky
We are developing a class of systems, called FAQ FINDER systems, that use a natural language questionbased interface to access distributed text information sources, specifically text files organized as question/ answer pairs such as FAQ files (Hammond, et al. 1995). In using these systems, a user enters a question in natural language and the system attempts to find an information source that answers the question, and then find the closest matching question/answer pair. As previous publications have described, these systems combine three technologies: statistically based IR engines, syntactic natural language analysis and semantic networks. In particular, they combine the SMART information retrieval system, a natural language parser, a semantic net derived from Princeton’s WOI~DNET, and question-answering strategies that handle questions that cannot be answered by straightforward question matching.