Data-driven Knowledge Engineering for CBR Systems

This keynote will focus on how data-driven methods facilitate the creation of case-based reasoning (CBR) systems. Traditionally, CBR systems are created in collaboration with domain experts who transfer their experience on comparing similar situations into cases and case representations, similarity measures, or adaptation knowledge. Data-driven methods can help to utilize existing datasets for creating the foundations of a CBR system. In this keynote, we will present (1) how machine learning methods can define case representations, (2) how similarity measures can be derived from data, and (3) how domain experts can be involved in the development process. We will use healthcare and aquaculture projects examples to showcase how these methods can be implemented using the open-source tool myCBR.