Call for Late-Breaking Results

We are pleased to invite you to contribute to the program of AIED2023 by submitting your late breaking results. The late-breaking results track offers an opportunity for presenting compelling, preliminary results and innovative work in progress. The goal is to give new, but not necessarily mature work a chance to be seen by other researchers and practitioners and to be discussed at the conference. Accepted submissions will be presented during the conference as posters.

The 24th international conference on Artificial Intelligence in Education (AIED) will take place between 3-7 July, 2023 in Tokyo, Japan and virtually. Its theme will be:AI in Education for Sustainable Society

The conference will be the latest of a longstanding series of international conferences, known for high quality and innovative research on intelligent systems and cognitive science approaches for educational computing applications. To celebrate the 30th anniversary of the AIED Society, we invite papers exploring how researchers envision the way AIED can shape the future of education in the next 30 years. AIED 2023 solicits empirical and theoretical papers particularly (but not exclusively) in the following lines of research and application:

  • AI-assisted and Interactive Technologies in an Educational Context;
  • Modeling and Representation;
  • Models of Teaching and Learning;
  • Learning Contexts and Informal Learning;
  • Evaluation;
  • Innovative Applications;
  • Equity and Inclusion in Education;
  • Ethics and AI in Education;
  • Explore Design, Use, and Evaluation of Human-AI Hybrid Systems for Learning; and
  • Online Learning Spaces.

Please see the main call for details about each of these topics


The AIED Society values diversity, equity, and inclusion (and related principles under this broad umbrella) as essential and fundamental values for the AIED community to uphold. Thus, in AIED 2023, we incentivize authors to carefully consider diversity, equity, and inclusion when reporting on your work. Please see the submission instructions for specific considerations.


All submissions must be in Springer format. Papers that do not use the required format may be rejected without review. Authors should consult Springer’s authors’ guidelines and use their proceedings templates, either for LaTeX or for Word, for the preparation of their papers. Springer encourages authors to include their ORCIDs in their papers. Submissions are handled via EasyChair:

Accepted papers for the Late-Breaking Results track will be published by Springer Communications in Computer and Information Science (CCIS).

Maximum paper length is as follows:

  • Late-breaking results papers (4 pages including references; will be presented as a poster)

Each accepted paper will be expected to have at least one author registered to attend in-person who will present the poster at the conference.

All submissions will be reviewed by the program committee to meet rigorous academic standards of publication. The review process will be double-masked, meaning that both the authors and reviewers will remain anonymous. To this end, authors should: (a) eliminate all information that could lead to their identification (names, contact information, affiliations, patents, names of approaches, frameworks, projects and/or systems); (b) cite own prior work (if needed) in the third person; and (c) eliminate acknowledgments and references to funding sources. Papers will be reviewed for relevance, novelty, technical soundness, significance and clarity of presentation. It is important to note that the work presented should not have been published previously or be under consideration in other conferences of journals. Any paper caught in double submission will be rejected without review.


  • Late-breaking results submission: March 6, 2023
  • Notification of decisions: April 10, 2023
  • Camera-ready version: May 1, 2023


When preparing your paper, please consider the following:

(1) Authors should write with care toward inclusive language. This includes understanding identify-first vs. person-first language, gender neutral language, appropriate demographic categories and terminology, and avoiding the conflation of distinct dimensions such as race and ethnicity, or sex and gender.

(2) Authors are encouraged to consider how their theoretical frameworks and findings are related to diversity, equity, and inclusion. For example, authors may discuss how these issues influence key assumptions, hypotheses, and methods. Likewise, authors might address implications or appropriate interpretations of their findings with respect to diversity, inclusion and equity.

Please consider the following criteria when reporting samples:

(1) Authors should be clear and specific about the composition of human-sourced data. Who were the participants? What was the distribution of gender, race, ethnicity, or related variables? If corpus data or training data were sourced from humans, a similar description could be offered.

(2) Skewed or non-representative samples would not necessarily trigger a "reject" decision, but authors should acknowledge the demographic imbalances and discuss the potential impact on data, results, or conclusions. A more compelling paper would describe barriers to inclusive and representative sampling and the steps taken to generate an inclusive and representative sample (this is basic science, but often overlooked for convenience).

(3) Authors should demonstrate some awareness of how equity, inclusion, accessibility issues impact their data, methods, products, or findings. How are different demographic groups or communities differentially connected to the work? People who are developing educational technologies need to think about access and use, for example. Corpus analyses need to address the impact of skewed/exclusive datasets and potential outcomes (e.g., algorithmic bias). It is also important to use strategies to control or reduce bias against populations of any kind (e.g., benefit or bring prejudice to a particular gender, race, or people with different economic status) when collecting, using, or aggregating data.

(4) Authors are encouraged to discuss/justify how demographic variables are included in the analyses. If they are not included or "covaried out" please justify. If they are included, what are the assumptions? Are there "categorical effects"? Are the effects of different demographic variables independent, interdependent, or intersectional? What valid conclusions can be drawn? What erroneous conclusions need to be avoided or tempered?


General Chairs

  • Noboru Matsuda, North Carolina State University, USA
  • Vania Dimitrova, University of Leeds, UK
  • Olga C. Santos, UNED, Spain

Program Co-chairs

  • Ning Wang, University of Southern California, USA
  • Genaro Rebolledo-Mendez, The University of British Columbia

Local Chair

  • Maomi Ueno, University of Electro-Communications, Japan

Posters and Late-Breaking Results Co-chairs

  • Carrie Demmans Epp, University of Alberta, Canada
  • Marie-Luce Bourguet, Queen Mary University of London, UK
  • Andrew M. Olney, University of Memphis, USA