Explanation Generation in Expert Systems (A Literature Review and Implementation)
Abstract
Today's technology provides tremendous amounts of information at incredible speeds. In order to make this information useful for more complex, significant problem solving applications, intelligent computer software systems are needed. The Expert System (ES) technology of Artificial Intelligence (AI) is one solution that is emerging to meet this need. However, as this technology continues to develop and we begin to use expert machines more and more, it is crucial that we demand the same explanatory capability from these mechanical experts as we do from human experts. The credibility of human expertise is established by an expert's ability to explain his expertise. The credibility of mechanical expertise must be established in the same way. Additionally, ESs are very complex, sophisticated systems. In order to verify their accuracy and correctness, ESs must be able to explain what they are doing and why. This thesis examines the Explanation Facilities (EFs) of ESs by first conducting an extensive literature review of this topic, and second, by implementing an EF for a frame-based ES shell. The purpose of the literature review is to gain a general understanding of EF research and development while the purpose of the implementation effort is to investigate the specifics of: explaining a frame- based ES, adding an EF to an existing ES, and using Ada as the implementation language for this AI application. Theses. (EDC)
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 01, 1989
- Accession Number
- ADA218741
Entities
Organizations
- Air Force Institute of Technology