David W. Aha and Ashwin Ram
Our motivation for organizing the 1995 AAAI Fall Symposium on Adaptation of Knowledge for Reuse derives from our shared concerns regarding case-based reasoning (CBR) research and commercial tools. CBR algorithms differ from similar algorithms studied in related fields (e.g., information retrieval, planning, machine learning, pattern recognition) in that they explicitly adapt retrieved solutions to solve new problems. Adaptation is an exciting topic in CBR; several researchers have proposed adaptation algorithms. Yet David Leake’s summary of the AAAI-93 Workshop on CBR confirmed what we suspected: while commercial CBR tools can efficiently retrieve cases, none include methods for automatically adapting them (i.e., users are limited to coding domain-specific adaptation rules). This was further clarified at the more recent CBR Workshops (e.g., at AAAI-94, EWCBR-95, and IJCAI-95), where both researchers and practitioners lamented this absence.