Argumentation and Machine Learning: When the Whole is Greater than the Sum of its Parts
Date: 12 August 2019, morning session
Lecturers: Federico Cerutti
Location: Macao, China, The Venetian Macao Resort Hotel, Naples 2703-2704
When: Mon, August 12 2019, morning session.
Where: Macao, China, The Venetian Macao Resort Hotel, Naples 2703-2704.
Who: Federico Cerutti.
Recording
Short Description of the Tutorial
Argumentation technology is a rich interdisciplinary area of research that, in the last two decades, has emerged as one of the most promising paradigms for commonsense reasoning and conflict resolution in a great variety of domains, to the point that it is used in commercial research projects such as IBM Debater.
In this tutorial PhD students, early-stage researchers, and machine learning experts are introduced to argumentation technology with concrete practical examples, state-of-the-art argumentation approaches that leverage machine learning, and state-of-the-art machine learning approaches that leverage argumentation technology.
Prior knowledge of argumentation theory is not required.
Description
The tutorial first explores the elements underpinning most approaches in argumentation theory and highlights the connections among epistemology, law, and complexity theory. It then reviews research-grade prototypes that apply argumentation to real problems, extending the scope of applications from legal reasoning to sense-making in intelligence analysis.
The second part discusses how machine learning can help with knowledge acquisition and with identifying the most suitable algorithms for argumentative reasoning. This includes argument mining, which sits at the intersection of argumentation and natural language processing, as well as investigations into the use of machine learning techniques to improve the performance of solvers for computing semantics extensions of argumentation frameworks.
The final part reviews current approaches that embed argumentation in their learning architecture, sometimes as a regulariser and often as a support for explainability and algorithmic accountability. The tutorial is designed to leave attendees with a comprehensive understanding of both the technological state of the art in argumentation and the synergies already emerging with machine learning.