- When: 13 September 2022
- Where: Cardiff, Wales, UK
- Submission Deadline (final extension) : 23:59 (BST) 17 July 2022
- Notification Due: 8 August 2022
The significant strides made by subsymbolic machine learning (ML) methods in recent times has led some to adopt the perspective that symbolic methods, or “good old-fashioned AI”, are becomming less relevant. However, there is a growing interest in synergistic AI systems that incorporate both subsymbolic and symbolic processes. Argumentation, as a symbolic model for effective human reasoning, offers many theoretical and practical opportunities for developing effective synergy. For instance, Argumentation Theory can substantially enhance the application of ML methods to various domains – such as discourse analysis, law, forensic analysis, to name just a few – by constraining the derivation of a solution to follow a justifiable and auditable process, in accordance with expert domain knowledge. The desire to build explainable AI (XAI) systems, that are tractable to human understanding and interaction, also presents an intuitive and fertile ground for argumentation-based systems to have a significant role. In turn, the challenges of establishing effective argumentation-based systems – such as (but not limited to) costly reliance on expert judgement, noise intolerance and brittle models, knowledge acquisition – may be significantly ameliorated by successfully embracing data-driven ML methods. Embracing ML methods presents an opportunity to draw from their effective use of computational resources, with the potential to vastly increase the scalability of argumentation-based methods in the age of big data. This workshop solicits contributions to meet these aims.
List of Topics
- Machine Learning-driven Reasoning Algorithms
- Neural-symbolic Learning
- Explainable AI
- Machine Learning and Argumentation for Agents and Multi-agent Systems
- Learning Symbolic Abstractions from Unstructured Data
- Machine Learning Applications in Argumentation
- Knowledge-driven Decision Making
- Architectures that Combine Data-driven Techniques and Formal Reasoning
- Expressive Power of Learning Representations
- Machine Learning for Efficient Knowledge Inference
- Learning Causal Models
- Machine Learning and Argumentation in Robotics
- Applications of Argumentation and Machine Learning in Law, Medicine, Finance, Policy Security, etc.
- Full papers: 12 pages + references
- Short/Position papers: 6 pages + references
- Submissions are NOT anonymous. The names and affiliations of the authors should be stated in the manuscript.
- All papers must be original and not simultaneously submitted to another journal or conference.
- All papers should be formatted following the IOS Press style and submitted through the EasyChair link accessed via the submission button at the bottom of the page.
- At least one of the authors will be required to register and attend the COMMA conference to present the paper in order for it to be included in the workshop proceedings.
ArgML papers will be published in CEUR-WS proceedings.
All questions about submissions should be emailed to Jack Mumford jack[dot]mumford[at]liverpool.ac.uk