News | Journal of Artificial Intelligence Research

2024 IJCAI-JAIR Prize Awarded

We congratulate Curtis Northcutt, Lu Jiang, and Isaac Chuang, who have been awarded the 2024 IJCAI-JAIR best paper prize for their article "Confident Learning: Estimating Uncertainty in Dataset Labels", which appeared in JAIR Volume 70 (2021).

As noted in the award citation, the work focuses on data label noise, providing both theoretical and empirical perspectives on how to handle this pressing issue. A strong contribution is a method for estimating the joint distribution of noisy and clean labels without regard to data modality, leading to actionable insights that are grounded in theory. The paper forms the basis for a widely-used open-source software library that has led to significant practical impact.

The IJCAI-JAIR Best Paper Prize is one of the most prestigious awards for a single publication in the field of AI. It has been awarded annually since 2003 to an outstanding paper published in JAIR in the last 5 calendar years, selected based on the significance of the work and the quality of the presentation.

This year’s selection committee consisted of Matthijs Spaan (TU Delft, chair), Haris Aziz (University of New South Wales), Alan Fern (Oregon State University), Julia Hockenmaier (University of Illinois at Urbana-Champaign), Akshat Kumar (Singapore Management University), and Gilles Pesant (Polytechnique Montréal).

Special Track on Multi-Agent Path Finding

JAIR invites submissions of original research for consideration for our new special track on Multi-Agent Path Finding (MAPF). MAPF is the abstract combinatorial problem of computing collision-free movement plans for a team of cooperative agents. The ability to solve instances of MAPF, efficiently and effectively, is a key enabler for many current and emerging industrial applications. These include warehouse logistics, train planning, pipe routing, robotic manufacturing, and many others. For more details see the special track page at: https://www.jair.org/index.php/jair/SpecialTrack-MAPF.

Special Track Editors: Daniel Harabor, Monash University; Sven Koenig, University of California Irvine; Jingjin Yu, Rutgers University

Submission period: July 1, 2024 - December 31, 2024

Target date of completion for handling the submissions (including resubmissions): December 2025

Special Track on Fairness and Bias in AI

JAIR invites submissions of original research for consideration for our special track on Fairness and Bias in AI. AI-based decision support systems are increasingly deployed across out societies to guide decisions in important spheres, including hiring decisions, university admissions, loan granting, medical diagnosis, and crime prediction. However, we lack a comprehensive understanding of how concepts of bias or discrimination should be interpreted in the context of AI and which socio-technical options to combat bias and discrimination are both realistically possible and normatively justified. For more details and a non-exhaustive list of topics see the special track page at: https://www.jair.org/index.php/jair/SpecialTrack-FBAI.

Special Track Editors: Roberta Calegari, University of Bologna; Andrea Aler Tubella, Umeå University; Virginia Dignum, Umeå University; Michela Milano, University of Bologna

Submission period: December 1st, 2023 - July 1, 2024

Target date of completion for handling the submissions (including resubmissions): July 2025

Dragomir R. Radev, Remembered

By Min-Yen Kan

Professor Dragomir R. Radev of Yale University, passed on March 29, 2023. He was 54. He was the surveys editor for our Journal of Artificial Intelligence Research for over nine years, directing the reviewing, editing and publishing of over 60 JAIR surveys in his stewardship. He was a distinguished member of the computing community, earning broad recognition as an educator and researcher, as a fellow of the AAAI, AAAS, ACM and the ACL.

In many ways, I have been following in Drago’s footsteps for many years. I first met him when I was an undergraduate student researcher at Columbia. He was a hearty man, epitomizing the phrase “jolly good fellow”, with his warm demeanor, strong sense of duty and responsibility for the next generation. I know many members of the natural language processing and computational linguistics communities had the privilege of benefiting from Drago tutelage in the early stages of their careers, as he was one of the drivers of the North American Computational Linguistics Olympiad (NACLO) and served as the secretariat of the ACL. I followed him in becoming a student of our supervisor Kathy, becoming the webmaster of our Columbia Computer Science department (when the Web was a tiny thing), taking on faculty responsibilities as an assistant professor, and serving our natural language processing and computational linguistics communities as a leader within the ACL. Even in research, we shared mutual interests, in bridging both natural language processing and information retrieval communities and bettering scholarly communication.

His dedication to others in his work was certainly mirrored in his family life too. There are many elements of his life story that have been shaped by personal circumstances. To some, such circumstances may be perceived as setbacks, but I believe Drago’s personality and his successes also stem from his perseverance that comes from his unwavering moral compass to do good and serve others.

Early February this year, I had my last conversation with Drago while he was on the move between work and personal appointments in New York City. He asked me to consider being his successor. He had told me he was to step down from JAIR’s survey editor position as he had been coordinating the surveys for a number of years and that his other duties had continually increased in load. True to form, Drago also took the opportunity to mentor: he shared with me how much he had learned from his JAIR duties, how beneficial it would be to know AI research more broadly through the editorship, and scheduled time to walk me through the process.

I owe Drago a huge debt in his mentorship and role model for how upright faculty should conduct themselves in both personal and professional roles. I am honored to follow in Drago’s footsteps one last time. Rest in peace, Drago!

2023 IJCAI-JAIR Prize Awarded

We congratulate Rodrigo Toro Icarte, Toryn Klassen, Richard Valenzano, and Sheila McIlraith, who have been awarded the 2023 IJCAI-JAIR best paper prize for their article "Reward machines: Exploiting reward function structure in reinforcement learning”, which appeared in JAIR Volume 73 (2022).

As noted in the award citation, This work considers how the structure of an explicitly-described reward function can be exploited to support efficient learning. This important setting has so far been under-explored; reward functions are typically expert-designed and therefore contain structure that most existing work ignores. The article proposes representing such a structure using a reward machine, a richly expressive type of finite-state machine; it also describes several powerful and effective methodologies to exploit the resulting structure. This article is the culmination of an influential line of research that bridges and enriches three different communities: AI planning, reinforcement learning, and formal methods.

This year’s selection committee consisted of Mykel Kochenderfer, (Stanford University, chair), George Konidaris (Brown University), Renata Wassermann (University of Sao Paulo), Matthijs Spaan (TU Delft) and Nick Hawes (Oxford University).

JAIR Now Available in ACM Library

JAIR and ACM are very pleased to announce that JAIR articles are now being hosted by the ACM Digital Library, in addition to the JAIR.org website. While the journal will continue to be managed and published as an independent, open access journal, this partnership will provide greater visibility for JAIR articles and their authors.

JAIR Transparent Publishing

When it was launched in 1993, JAIR was one of the very first open access scientific journals. Since then, JAIR has not only emerged as one of the top publication venues in artificial intelligence, but also inspired the creation of other, similarly successful open access journals.

Now, JAIR is taking another major step by launching its transparent publishing initiative. Starting today, JAIR publishes, on a regular basis, detailed metrics on its submission handling process, including empirical likelihoods for all evaluation outcomes and average times for reaching these outcomes. This transparent publishing approach is intended to provide useful information to prospective authors and valuable calibration to the editorial team and to reviewers.

More information about JAIR Transparent Publishing and the latest metrics are found here.

Please Support JAIR

AI Access Foundation manages JAIR largely through the efforts of volunteers throughout the world. However, we occasionally have some small, but important, operating and infrastructure costs. Please support us by making a donation. All donations are appreciated, no matter how small, and they are tax dedectable.

And on this topic, we also want to thank David Smith for his financial contributions to the organization, as well as IJCAI, and InferLink Corporation for their ongoing support for JAIR's infrastructure. These gifts make it possible for JAIR to continue operating a freely-available journal.