Kind regards,

Patrick Mannion

Fernando P. Santos

Diederik M. Roijers

—

As autonomous agent-based systems become ever more prevalent in everyday life, it is imperative that society can trust that such systems will act for the benefit of humanity. Ensuring trustworthiness for autonomous systems is one of the key global challenges facing society at present, as evidenced by recently published guidelines on the topic by organisations such as the European Commission, the IEEE, and the OECD. Trustworthiness has a number of different dimensions, including explainability, safety, fairness, accountability and compliance with legislative and ethical standards.

Autonomous agents operating in the real world should therefore make decisions in a fair and transparent manner that respects ethical principles, should be aware of their social environment and should comply with applicable regulations. This can prove challenging given the complexity of agent architectures and the long-term dynamics — often hard to anticipate and control — resulting from multiple agents learning and adapting to each other and to constantly changing environments. Furthermore, the majority of published research on autonomous agents does not explicitly consider the level of trustworthiness of the proposed approaches, leaving a vast gap in the literature between the theory and practical application of agent-based systems.

Learning and adaptation are key capabilities for autonomous systems. This topical collection (TC) in the AI and Ethics (AI&E) journal focuses on the topic of Trustworthy Adaptive and Learning Agents (TALA). AI&E is a new journal recently launched by Springer, and seeks to promote informed debate and discussion of the ethical, regulatory, and policy implications that arise from the development of AI. The TALA TC targets high-quality original papers covering all aspects of trustworthiness in agent-based systems, including, but not limited to, the list of topics below. Manuscripts that extend a previous conference or workshop publication are welcome, provided that there is a significant amount of new material in the submission (i.e., the manuscript should contain at least 30% new material).

This topical collection is associated with the long-running and successful series of workshops on Adaptive and Learning Agents (ALA), that have been held each year since 2009 in conjunction with the AAMAS conference. Therefore, manuscripts reporting extended versions of work presented at a prior edition of the ALA workshop are very much welcome. The TALA TC has an open call for papers; it is not necessary to submit preliminary work to the ALA workshop in order to have your manuscript considered for publication in this TC.

The following is a non-exhaustive list of topics that we would like to cover in the special issue:

Trustworthy algorithms for ALA, including those based on reinforcement learning and planning

Principled approaches to reward design for trustworthy ALA

Trustworthy multi-agent decision making

Requirements and design principles for trustworthy ALA 

Benchmark problems for verifying trustworthiness of ALA 

Multi-objective decision making approaches to TALA 

Analyses of TALA from different ethical paradigms (such as utilitarianism, deontology, particularism, etc.).

Handling (environmental epistemic and aleatoric) uncertainty in TALA 

Safe reinforcement learning

Explainable (learning) agents

Avoidance of bias in ALA

Emergence of coordination among adaptive and learning agents towards societal and environmental well-being

Long-term trustworthiness in dynamic environments composed of learning agents

Game theoretic approaches to frame ethical dilemmas in multiagent systems

Agent-based approaches to model the societal impacts of AI 

Compliance of ALA with regulations, ethics and/or social norms

Methods to counter malicious effects of autonomous agents (e.g., preventing misinformation through bots on social media)

Perspectives on cultural differences in accepting and trusting autonomous learning agents 

Approaches to audit the behavior and impact of ALA, including agent failures

Patrick Mannion (Lead Guest Editor), School of Computer Science, National University of Ireland Galway, webpage, email: patrick.mannion@nuigalway.ie 

Fernando P. Santos, University of Amsterdam, webpage, email: f.p.santos@uva.nl 

Diederik M. Roijers, Vrije Universiteit Brussel & HU University of Applied Sciences Utrecht, webpage, email: diederik.roijers@vub.be 

There is no specific submission deadline for this TC. Manuscript submissions will be considered for publication in the TALA TC on a continuous basis until a sufficient number of manuscripts have been accepted for publication. Manuscripts will be sent out for review as soon as they are received, and first decisions on manuscripts can be expected within 2 months approx. from the initial submission date. Submissions accepted for publication before the completion of the topical collection will be published online on the journal website shortly after acceptance. Authors considering submitting to the TALA TC should contact the Guest Editors in advance, to ensure that their proposed manuscript is in scope, and that there is space in the TC for the manuscript.

This TC solicits original research articles, reviews/surveys, and opinion pieces/commentaries relating to trustworthiness in agent-based systems, including those that employ learning and/or adaptation. Research articles should present original and high-quality theoretical and/or empirical results that advance the field of Trustworthy Adaptive and Learning Agents. It is expected that original research articles include (as appropriate) full Introduction, Background, Related Work, Methods, Results, and Discussion sections. Reviews/surveys should provide a comprehensive summary of a research topic of interest to TALA, and identify open challenges and new research directions for the field based on a thorough analysis of current literature. Opinion pieces/commentaries should offer new personal perspectives, visionary ideas, current challenges or summarize new research opportunities on a topic related to TALA, be circa 2500-5000 words and be accessible to a broad scientific audience.

Before submitting, authors should read the AI&E submission guidelines at https://www.springer.com/journal/43681 in full. To submit, you should visit the online system at https://www.editorialmanager.com/aiet and create a new author account if you do not already have one. When creating your submission on the system, select the article type (e.g., Original Research, Review, or Opinion Paper), and then in the “Additional Information” section, answer “Yes” when asked if your manuscript belongs to a special issue, then select “T.C. : Trustworthy Adaptive and Learning Agents TALA”. If you do not mark your manuscript correctly as belonging to the TALA topical collection, it may not reach the correct editors.

Deadline Extended: Adaptive and Learning Agents Workshop at AAMAS 2020 (Auckland, New Zealand)

** Apologies if you receive more than one copy. Please share with students and colleagues. **

Dear all,

We are organizing the next iteration of the Adaptive and Learning Agents (ALA) workshop at the International Conference on Autonomous Agents and Multiagent Systems (AAMAS) in Auckland. Please find the CfP below.

*******************************************************

Adaptive and Learning Agents Workshop at AAMAS (Auckland, New Zealand)

Extended submission deadline: February 24, 2020

Extended versions of all original contributions at ALA 2020 will be eligible for inclusion in a special issue of the Springer journal Neural Computing and Applications (Impact Factor 4.213).

*******************************************************

TL;DR:

* Workshop with a long and successful history, now in its twelfth edition.

* Covering all aspects of adaptive and learning agents and multi-agent systems research.

* Open to original research papers, work-in-progress, and visionary outlook papers, as well as presentations on recently published journal papers.

* ACM proceedings (AAMAS) format up to 8 pages (excluding references) for original research, up to 6 pages for work-in-progress and outlook papers (shorter papers are also welcome and will not be judged differently) and 2 pages for recently published journal papers.

* Accepted papers are eligible for inclusion in a post-proceedings journal special issue.

* Submissions through easychair: https://easychair.org/conferences/?conf=ala2020

*******************************************************

IMPORTANT DATES:

* Submission Deadline:          February 24, 2020

* Notification of acceptance:  March 10, 2020

* Camera-ready copies:         March 24, 2020

* Workshop:                           May 9 & 10, 2020

*******************************************************

OVERVIEW

Adaptive and learning agents, particularly those interacting with each other in a multi-agent setting, are becoming increasingly prominent as the size and complexity of real-world systems grows. How to adaptively control, coordinate and optimize such systems is an emerging multi-disciplinary research area at the intersection of Computer Science, Control Theory, Economics, and Biology. The ALA workshop will focus on agents and multi-agent systems which employ learning or adaptation.

The goal of this workshop is to increase awareness of and interest in adaptive agent research, encourage collaboration and give a representative overview of current research in the area of adaptive and learning agents and multi-agent systems. It aims at bringing together not only scientists from different areas of computer science but also from different fields studying similar concepts (e.g., game theory, bio-inspired control, mechanism design).

This workshop will focus on all aspects of adaptive and learning agents and multi-agent systems with a particular emphasis on how to modify established learning techniques and/or create new learning paradigms to address the many challenges presented by complex real-world problems.

The topics of interest include but are not limited to:

   * Novel combinations of reinforcement and supervised learning approaches

   * Integrated learning approaches that work with other agent reasoning modules like negotiation, trust models, coordination, etc.

   * Supervised multi-agent learning

   * Reinforcement learning (single and multi-agent)

   * Novel deep learning approaches for adaptive single and multi-agent systems

   * Multi-objective optimisation in single- and multi-agent systems

   * Planning (single and multi-agent)

   * Reasoning (single and multi-agent)

   * Distributed learning

   * Adaptation and learning in dynamic environments

   * Evolution of agents in complex environments

   * Co-evolution of agents in a multi-agent setting

   * Cooperative exploration and learning to cooperate and collaborate

   * Learning trust and reputation

   * Communication restrictions and their impact on multi-agent coordination

   * Design of reward structure and fitness measures for coordination

   * Scaling learning techniques to large systems of learning and adaptive agents

   * Emergent behavior in adaptive multi-agent systems

   * Game theoretical analysis of adaptive multi-agent systems

   * Neuro-control in multi-agent systems

   * Bio-inspired multi-agent systems

   * Adaptive and learning agents for multi-objective decision making

   * Applications of adaptive and learning agents and multi-agent systems to real world complex systems

*******************************************************

SUBMISSION DETAILS

Papers can be submitted through EasyChair: https://easychair.org/conferences/?conf=ala2020

We invite submission of original work, up to 8 pages in length (excluding references) in the ACM proceedings format (i.e. following the AAMAS formatting instructions). This includes work that has been accepted as a poster/extended abstract at the AAMAS 2020 conference (we encourage authors to append the received reviews and a short letter highlighting the changes carried out on the paper; the reviews can be appended at the end of the submission file and do not count towards the page limit). Additionally, we welcome submissions of preliminary results, i.e. work-in-progress, as well as visionary outlook papers that lay out directions for future research in a specific area, both up to 6 pages in length, although shorter papers are very much welcome, and will not be judged differently. Finally, we also accept recently published journal papers in the form of a 2 page abstract.

All submissions will be peer-reviewed (single-blind). Accepted work will be allocated time for poster and possibly oral presentation during the workshop. Extended versions of all original contributions at ALA 2020 will be eligible for inclusion in a special issue of the Springer journal Neural Computing and Applications (Impact Factor 4.213). Deadline for submitting extended papers: September 15, 2020.

We look forward to receiving your submissions,

– The Organizers

Patrick MacAlpine (Microsoft Research, USA)

Fernando P. Santos (Princeton University, USA)

Felipe Leno da Silva (University of São Paulo, BR)

Roxana Rădulescu (Vrije Universiteit Brussel, BE)

==================================================================