VALIDATION OF KNOWLEDGE-BASED SYSTEM WITH MULTIPLE BRIDGE RAIL EXPERTS

Public-works projects usually undergo review by multiple experts at various levels. This technical paper describes a validation procedure for rule-based expert systems. Four qualitative measures are applied, which are based on knowledge extracted from experts working at various hierarchical levels. Those measures are face validation, field test, Turing test, and sensitivity analysis. The authors apply the proposed method to a system that ranks bridge rails from a list of choice configurations. The system aids engineers in selecting bridge rails according to road, bridge, and environmental conditions. Based on the level of agreement with human experts and the Turing test, the prototype system demonstrates a high level of expertise. Some system deficiencies exist and several features need refinement.

Language

  • English

Media Info

Subject/Index Terms

Filing Info

  • Accession Number: 00669221
  • Record Type: Publication
  • Files: TRIS, ATRI
  • Created Date: Oct 31 1994 12:00AM