KAVAS-2: knowledge acquisition, visualisation and assessment system

The KAVAS-2 (A2019) AIM project develops a toolbox, named KAVIAR, with which the domain expert can formulate (capture) his/her own knowledge and/or derive domain models (so called `learning') on the basis of example cases on the clinical problem under focus. Key issues for the global learning process within KAVIAR are 1) quality assessment of the derived domain models, 2) controlling the domain model generation in a desired direction, and 3) coupling of machine learning techniques with knowledge elicitation techniques. Other main issues addressed in KAVAS-2 are the cognitive aspects in the display of information and knowledge, the refinement of an evaluation methodology for KBS, and elucidation of problems inherent in the transfer of medical data, information and knowledge and hence, of dissemination of medical KBS.