Integrated Case Based and Rule Based Reasoning for Decision Support
Azeb Bekele Eshete · 2009
Knowledge based decision support systems can benefit from combining generalized experiences, e.g. in the form of rules, with situation-specific experiences, in the form of past cases. In this thesis architecture for an integrated system, combining CBR and RBR methods, shall be developed, a demo system that instantiates essential parts of the architecture shall be designed, implemented, and tested. The rule based tool Jess shall be used for the rule based part, and a freely available tool shall be chosen for the case based part. The application domain in this study will be oil well drilling. The KBS group at IDI cooperates with the company Verdande Technology As, which builds software for the oil and gas industry in which case-based reasoning methods are used. For this thesis a sub-problem of oil drilling has to be selected. Cases should be constructed and represented within the chosen CBR tool, partly based on existing cases in Verdande Technology’s case base. Rules should be acquired from documents that contain generalized experience in the form of best practice and lessons learned.