Analysis of Representational Design Decisions to Create Ontologies and Guidelines

• VANWELKENHUYSEN Johan

  I. S. I. R., Osaka University

• MIZOGUCHI Riichiro

  I. S. I. R., Osaka University

抄録

We have analysed domain models about processor boards of an existing diagnostic knowledge system to create sharable ontologies and allow for reuse of past expertise analyses and system design experiences. Design decisions and rationales behind representations in the domain models are uncovered to explain their adequacy for automated troubleshooting. This information is used to formulate guidelines which capture knowhow to refine or change representations to compromise between competing non-functional requirements and aspects of the problem solving task. Purpose of these guidelines is to facilitate the reverse construction from generic ontologies to refined concepts and relationships tuned to certain task environments. These guidelines add a novel element to the technology for knowledge sharing and reuse.

収録刊行物

• 人工知能学会誌

  人工知能学会誌 11 (2), 310-320, 1996-03-01

  社団法人人工知能学会

キーワード

• problem solving ontologies
• design rationales
• reuse
• diagnosis
• digital systems

詳細情報

• CRID: 1572543027054731648
• NII論文ID: 110002808626
• NII書誌ID: AN10067140
• ISSN: 09128085
• 本文言語コード: en
• データソース種別:
  • CiNii Articles