Track:
Contents
Downloads:
Abstract:
A robust information extraction system, which can accurately and rapidly convert vast amounts of textual data into a semantic network type knowledge representation scheme, will be a useful preprocessor for various link analysis applications. In this paper, we describe an end-to-end knowledge discovery system which takes raw texts as input and generates a knowledge base which can be utilized for a variety ofinforrnation access systems, such as a visual knowledge base browser and a question-answering system.