Machine Arguing

Module aims

In this module you will have the opportunity to:

  • Work on the active AI research area of Argumentation
  • Gain familiarity with foundational and advanced concepts in Machine Aruging
  • Understand how Machine Arguing exists within the context of several application domains and is at the intersection with knowledge representation and reasoning, multi-agent systems, natural language processing
  • Contribute formalisms, systems and applications for settings where the resolution of conflicts, within and across entities, is key and where explanation of how these conflicts are resolved is essential.

Machine Arguing amounts to

  • the definition and study of argumentation frameworks (i.e. formalisms for modelling conflicts)
  • the definition, study and implementation of (dialectical or gradual) semantics and algorithms  for these argumentation frameworks  (i.e. methods for resolving conflicts)
  • the definition and implementation of methods for mining argumentation frameworks from a variety of sources, including data of various types and logical rules

Learning outcomes

Upon successful completion of this module you will be able to:

  • explain the foundational concepts and methods of Machine Arguing
  • critically compare and evaluate the methods of Machine Arguing
  • develop and reason about applications of Machine Arguing

Module syllabus

Argumentation frameworks, semantics, algorithms:

  • abstract argumentation
  • assumption-based argumentation
  • bipolar (and Tripolar) argumentation
  • argumentation and preferences
  • argumentation and probabilities

Mining argumentation frameworks:

  • argument mining in natural language processing
  • mining argumentation frameworks from labelled, feature-engineered data
  • mining argumentation frameworks from logical rules and from games
  • mining argumentation frameworks from recommender systems

Teaching methods

The course will be delivered by means of interactive lectures, whereby the lecturer will provide defininitions and illustratoins and you and the lecturer will engage in individual or group problem solving to firm understanding. The lecture notes will be integrated by tutorial material, equipped with sample answers to guide students.

The Piazza Q&A web service will be used as an open online discussion forum for the module.

Assessments

There will be two courseworks that each contribute 10% of the mark for the module - a total of 20%. There will be a final written exam, which counts for the remaining 80% of the marks. 

You will receive individual feedback on both pieces of submitted coursework as well as class aggregated feedback during the lectures

Reading list

Core

Useful

  • Elements of argumentation

    Besnard, Philippe, 1958-

    Cambridge, MA. ; London : MIT Press

  • Argumentation in artificial intelligence

    Bench-Capon, T.J.M. ; Dunne, Paul E.

    Artificial Intelligence

Module leaders

Professor Francesca Toni