Using naturally occurring texts as a knowledge acquisition resource for knowledge base design: developing a knowledge base taxonomy on microprocessors

Abstract

Many artificial intelligence applications suffer severely from a bottleneck in acquiring domain information necessary to go beyond toy hand-built demonstrations to realistic applications. This project examines one approach to reducing that bottleneck by using automated and semi-automated techniques to analyze published domain-relevant material. A taxonomy of terms related to computers with an emphasis on microprocessors is developed and presented. The methods used are experimental and not yet fully validated, but are potentially of great use for extracting useful domain information from published material. Preliminary validation by comparison with a published taxonomy shows that these methods have produced a taxonomy which is better suited for the immediate use of this taxonomy.