Invited Speakers

Kerstin Bach
Norwegian University of Science and Technology

"Developing Digital Healthcare Interventions using Case-Based Reasoning"


Abstract: CBR has a long tradition in medical applications. While most applications focus on the recommendation and decision support for clinicians in specialized care, the selfBACK project has created an app for low back pain patients based on CBR to create self-management plans and automatically follow up on their progress. This talk presents the selfBACK approach that delivers customized healthcare and lifestyle recommendations in a primary care setting to supplement traditional care. We show how we build the CBR system with an interdisciplinary team and which lessons we learned on basing our work on CBR. The effectiveness of the final application has been evaluated in a randomized controlled trial in which a total of 461 patients participated. All participants were randomized to receive self-management support in addition to usual care (selfBACK) or usual care alone.

Further, we will discuss the challenges and opportunities of using CBR as a methodology when building novel applications in interdisciplinary teams. In the end, we will give an outlook towards recently started projects in which we build upon the work from the selfBACK project aiming to support primary care clinicians in treatment planning or facilitating journeys of patients with musculoskeletal disorders returning to work.

Biography: Kerstin’s research interests are methods for developing AI systems with a particular focus on CBR systems. While the application domains differ, we investigate how to make knowledge and experience available through intelligent systems. Moreover, how to build systems that support complex, knowledge-intensive decisions using heterogeneous data sources.

Kerstin’s is an associate professor at the Department of Computer Science at NTNU in Trondheim, Norway, where she is the deputy group leader of the Data and Artificial Intelligence group and a member of the Norwegian Open AI Lab. Since August 2021, she is the program manager of a newly established Center for Research-based Innovation (SFI) – NorwAI: Norwegian Research Center for AI Innovation and associated with the Norwegian Open AI Lab. Kerstin was awarded her PhD in Computer Science from the University of Hildesheim, Germany, working with Klaus-Dieter Althoff on knowledge acquisition techniques for CBR systems and the open-source tool myCBR.

Santiago Ontañón Villar
Google Research and Drexel University

"Compositional Generalization in Machine Learning and CBR"


Abstract: Although modern machine learning techniques can achieve state-of-the-art performance in many challenging natural language, computer vision or robotics tasks, they still lag behind humans in several aspects. Specifically, one of them is "compositional generalization", the ability to learn a set of basic primitives and combine them in more complex ways than those seen during training. Compositional generalization is a key aspect of natural language and many other tasks we might want machine learning models to learn. In this talk, we will introduce the challenge of compositional generalization, argue how it is one of the biggest open challenges in modern machine learning, and discuss potential avenues for future work in the context of machine learning and case-based reasoning.

Biography: Santiago Ontañón is a Research Scientist at Google Research and an Associate Professor at Drexel University. His research focuses on machine learning, AI, game AI and case-based reasoning, areas in which he has published over 200 papers. He obtained his PhD at the Artificial Intelligence Research Institute (IIIA) under the supervision of Enric Plaza. Before joining Google and Drexel, he held postdoctoral positions at Georgia Tech, the University of Barcelona and IIIA.

©2021 | 29th International Conference on Case-Based Reasoning . All Rights Reserved.