Improvement of online adaptation knowledge acquisition and reuse in case-based reasoning: Application to process engineering design

Abstract

Interactive adaptation method Online knowledge acquisition Failure diagnosis and repair a b s t r a c t Despite various publications in the area during the last few years, the adaptation step is still a crucial phase for a relevant and reasonable Case Based Reasoning system. Furthermore, the online acquisition of the new adaptation knowledge is of particular interest as it enables the progressive improvement of the system while reducing the knowledge engineering effort without constraints for the expert. Therefore this paper presents a new interactive method for adaptation knowledge elicitation, acquisition and reuse, thanks to a modification of the traditional CBR cycle. Moreover to improve adaptation knowledge reuse, a test procedure is also implemented to help the user in the adaptation step and its diagnosis during adaptation failure. A study on the quality and usefulness of the new knowledge acquired is also driven.As our Knowledge Based Systems (KBS) is more focused on preliminary design, and more particularly in the field of process engineering, we need to unify in the same method two types of knowledge: contextual and general. To realize this, this article proposes the integration of the Constraint Satisfaction Problem (based on general knowledge) approach into the Case Based Reasoning (based on contextual knowledge) process to improve the case representation and the adaptation of past experiences. To highlight its capability, the proposed approach is illustrated through a case study dedicated to the design of an industrial mixing device.

How to share and utilise expertise in a police forensic department through externalisation and mutualisation

“…Then we will define the best knowledge management strategy that could be applied within this service. Those methodologies were currently mainly used to guarantee companies competitivity (Gavrilova et Andreeva 2012) and could be diverse as supervision (Köpsén et Nyström 2012;, community of practice (Ackerman et al 2013;Chiem et al 2014;Doak et Assimakopoulos 2007), Case-based reasoning (Casey 2013;Roldan Reyes et al 2015;Hoffman 1987;Aamodt et Plaza 1994;Schank et Abelson 1977) or case library (Kolodner et al 2003;Kulasegaram et al 2017;Tawfik et al 2017;Tawfik 2017). Even if some kind of supervision and community of practice (exchange between more and less experienced investigators) is used within almost every forensic department through interraction between investigators they are not seen as a knowledge management tool and their use is rarelly formalised.…”

An adaptive clinical decision support system for serving the elderly with chronic diseases in healthcare industry

“…However, currently, many researchers focus only on the case retrieval processes rather than case adaptation processes in CBR. This results in high uncertainties in the case adaptation processes due to the modification of retrieved cases that relies on the experience of domain experts (Reyes, Negny, Robles, & Lann, 2015). The uncertainties may affect the analysis quality due to imprecision and conflicts in human decision making (Jabrouni, Kamsu-Foguem, Geneste, & Vaysse, 2013).…”

An effective similarity determination model for case-based reasoning in support of low-carbon product design

“…Rousu and Aarts 39 incorporated the adaptation cost into the evaluation indicators of the similarity and adaptability. Roldan Reyes et al 40 reported that the CSP and CBR depend on the general knowledge and contextual knowledge respectively, and they integrated these two methods to address adaptation problems in the process of engineering design. Qi et al 41 incorporated the support vector machine (SVM) and adaptability-related knowledge to obtain adaptable cases, the knowledge of adaptability contains the adaptability characteristics of the old cases and the guideline to the construction of the adaptability-involving SVM adaptation model with respect to retrieved cases.…”