Curious CBR: Developing Self-Aware Case-Based Systems

Humans have innate knowledge acquisition skills. Learning, while sometimes triggered by specific information needs, often results from general curiosity about one’s environment along with an understanding of one’s goals and of the limitations in one’s existing knowledge. In contrast, automated reasoners lack curiosity and are unaware of their limitations. In Case-Based Reasoning (CBR) learning traditionally takes place by retaining the reasoner’s direct experiences, both successful and unsuccessful. While this approach has been shown to be successful in many problem-solving scenarios, it represents a very restrictive learning environment.
This project will develop a richer, more ``human-like’’ learning environment for reasoners that simulates human curiosity with a proactive learning approach. CBR has the potential to provide such systems. But in order to be effective in a modern rapidly changing problem-solving environment, they require stronger learning models that incorporate more flexible approaches to embedding learnt knowledge into the various knowledge containers that support problem-solving.

Essential Background: Equivalent of 2.1 Honours Degree in Computing Science or a related discipline. Knowledge of AI, Data Science and/or Big Data analysis would be an advantage. Good programming skills are essential.

Informal enquiries should be addressed to Dr. Stewart Massie ([Email Address Removed])