Scaling up explanation generation: large-scale knowledge bases and empirical studies

  • Authors:

  • James C. Lester;Bruce W. Porter

  • Affiliations:

  • Department of Computer Science, North Carolina State University, Raleigh, NC;Department of Computer Sciences, The University of Texas at Austin, Austin, Texas

  • Venue:
  • AAAI'96 Proceedings of the thirteenth national conference on Artificial intelligence - Volume 1
  • Year:
  • 1996

Quantified Score

Hi-index 0.00

Visualization

Abstract

To explain complex phenomena, an explanation system must be able to select information from a formal representation of domain knowledge, organize the selected information into multisentential discourse plans, and realize the discourse plans in text. Although recellit years have witnessed significant progress in the development of sophisticated computational mechanisms for explanation, empirical results have been limited. This paper reports on a seven year effort to empirically study explanation generation from semantically rich, large-scale knowledge bases. We first describe Knight, a robust explanation system that constructs multi-sentential and multi-paragraph explanations from the Biology Knowledge Base, a large-scale knowledge base in the domain of botanical anatomy, physiology, and development. We then introduce the Two Panel evaluation methodology and describe how Knight's performance was assessed with this methodology in the most extensive empirical evaluation conducted on an explanation system. In this evaluation, Knight scored within "half a grade" of domain experts, and its performance exceeded that of one of the domain experts.