Integration of Case Based Retrieval with a Relational
Database System in Aircraft Technical Support

Jonathan RC Allen1, David WR Patterson, Maurice D Mulvenna and John G Hughes

Northern Ireland Knowledge Engineering Laboratory
Faculty of Informatics, University of Ulster at Jordanstown Newtownabbey
Northern Ireland, UK, BT37 0QB



Abstract. Case-Based Reasoning (CBR) is suited to problem solving in
domains where there are recurring problems. This paper describes the
development of a CBR system for use in such a domain, the Technical
Support department of an aircraft manufacturing company. The system uses
three types of indexing: knowledge-guided induction, inductive indexing
and nearest neighbour matching. The resultant system integrates case based
retrieval with a relational database system to provide a rich environment to
help manage the life cycle of a technical support query. In early tests with
the system, staff can discern if a new query is a recurring problem and has
been solved before or if it is a completely new unsolved technical query.
References

1.	Barletta R. An Introduction to Case-Based Reasoning, Al Expert 8, 1991.
2.	Harmon, P. Case-Based Reasoning 2, Intelligent Software Strategies, 7(11), 1991.
3.	Kriegsman, M., & Barletta, R. Building a Case-Based Help Application, IEEE
Expert, 8(6), 1993.
4.	Harmon, P. Case-Based Reasoning 3, Intelligent Software Strategies, 8(1), 1992.
5.	Forsyth, R. Expert Systems: Principles and Case Studies, ed. Forsyth R, Chapman
and Hall Ltd, (1989).
6.	Michalski, R., Carbonnel, J., & Mitchell, T. Machine Learning: An Artificial
Intelligence Approach, Tioga Publishing Corp, Palo alto, CA, 1983.
7.	Magaldi, R.V., Maitaining Aeroplanes in Time-Constrained Operational Situations
Using Case-Based Reasoning, EWCBR 1994 Chantilly, France.
