Nonmonotonic Reasoning, Argumentation and Machine Learning

Reference: P. Clark. Nonmonotonic reasoning, argumentation and machine learning. Technical Report TIMLG-38, Turing Institute, Glasgow, UK, June 1990.

Abstract: Machine learning and nonmonotonic reasoning are closely related, both concerned with making plausible as well as certain inferences based on available data. In this document a brief overview of different approaches to nonmonotonic reasoning is presented, and it is shown how the concept of argumentation systems arises. The relationship with machine learning work is also discussed. The document aims to highlight the links between nonmonotonic reasoning, argumentation and machine learning and as a result propose some potentially useful directions for new research.