Utilizing Repair Cases of Home Electrical Appliances

Satoshi HORI  Hiromitsu SUGIMATSU  Soshi FURUKAWA  Hirokazu TAKI  

IEICE TRANSACTIONS on Information and Systems   Vol.E82-D   No.12   pp.1612-1617
Publication Date: 1999/12/25
Online ISSN: 
Print ISSN: 0916-8532
Type of Manuscript: PAPER
Category: Artificial Intelligence and Cognitive Science
field service,  case-based reasoning,  diagnostic expert system,  knowledge sharing,  

Full Text: PDF(782.1KB)>>
Buy this Article

We have developed a diagnostic Case-Based Reasoning (CBR) system, Doctor, which infers possible defects in a home electrical appliance and lists up necessary service parts. The CBR is suitable to build a diagnostic system for the field service because the CBR imitates how experienced service technicians infer and is able to learn defect trends and novel repair cases from a service report database. In order to apply a CBR system to this real-world problem, Our system has the following new features: (1) Its CBR mechanism utilizes not only repair cases, but also diagnostic rules that are elicited from human experts so that accurate diagnosis can be achieved. (2) Its casebase maintenance mechanism updates the casebase and adapts it to the changing real world.