Key Facts
Tutorial proposal
deadline:
February 18, 2022 AoE
Outcome notification
date (expected):
March 14, 2022 AoE
Conference:
June 21-24 2022 (Seoul and online)

Submissions Closed.

Overview

ACM FAccT solicits proposals for tutorials to be presented at the 2022 Conference, which will be held both in-person and online. The in-person conference will take place in the COEX exhibition centre in Seoul, South Korea on June 21-24 2022. Tutorials will be run mainly in-person, and presenters are expected to attend the in-person conference if possible. There is a substantial travel fund available to support in person attendance. We also expect a significant online audience, so to enable online participation in the tutorials and a good experience for the online audience, in addition to their live presentation, tutorial presenters will be asked to pre-record a version of the tutorial, and attend a 30 or 60 minute live online Q&A during the conference (early in the morning or in the evening, to cater to global time zones).

Tutorial Types

Tutorials should address practical, technical, policy, regulatory, ethical, or societal issues related to FAccT for a broad audience. We are soliciting three types of tutorials for 2022: dialogue/translation tutorials, implications tutorials, and practice tutorials. We will give presenters either 45 or 90 minutes to present their tutorial.

The goal of tutorials, whether introducing methods, presenting a broad overview of an application domain, or a deep dive on a case study, is to educate and broaden the perspective of our interdisciplinary community. Thus it is especially important that tutorial speakers be excellent educators (whether or not by profession).

The tutorials track is curated: the Tutorial Chairs will select from the pool of proposals with the aim of ensuring broad coverage of disciplines and topics. We welcome perspectives from a variety of disciplines, including computer science, economics, law, philosophy, political science, communication, sociology, education, and social work. A goal of the tutorials track is that the diversity of perspectives represents the diversity of the populations impacted by this technology throughout the world.

To better understand what kinds of tutorials fit with the conference, you could check the list of accepted tutorials from previous years: 2021 2020.

Dialogue/Translation Tutorials

We are interested in tutorials that build bridges between disciplines, through a dialogue around a concept or tool, or translating concepts from one discipline for other disciplinary audiences. A ‘dialogue’ type tutorial could involve an interdisciplinary team of presenters, taking turns to explain their own disciplinary approach to some key concept, followed by a dialogue/discussion between the experts. For instance, a member of the team with computer science expertise might explain the relevant computer science concepts, and a member of the team with legal expertise might explain their perspective on related concepts, followed by a discussion of the key ideas and lessons that their disciplines can exchange. Alternatively, a ‘translation’ tutorial could explain key concepts from one discipline in a way that is practically useful for other disciplinary audiences.

These tutorials should be geared towards an interested audience, but should not assume more than a beginner’s familiarity with the topics. Dialogue/Translation tutorials should situate the topic and the role it plays and proceed to a detailed explanation of that specific topic.

Implications Tutorials

Implications tutorials should cover known legal, policy, or socio-economic effects of the use of algorithmic systems in society. These tutorials should emphasize “real-world” implications with known examples. For instance, an implications tutorial may focus on specific case studies, walking the audience through the likely or known causes and effects of a particular FAccT issue for specific individuals, communities, or society more broadly.

We particularly encourage submissions by human rights / civil rights experts, including (but not limited to) lawyers, policy advocates, civil society representatives, and others who work closely with individuals and communities affected by algorithmic systems and who can offer a more in-depth understanding of the processes around the use of these systems.

Practice Tutorials

We also welcome tutorials focusing on the application, use, and deployment of tools and frameworks proposed by Fairness, Accountability, and Transparency scholarship in real-world practice. (The tools and frameworks at the center of such tutorials need not be limited to those offered by prior publications of the FAccT conference, but they should belong to the broader literature around issues of Fairness, Accountability, and Transparency for algorithmic, data-driven systems.) Tutorials should focus on tools and frameworks that have already been deployed/evaluated in actual practice or in the wild, not merely at a ‘proof-of-concept’ stage.

We particularly encourage submissions by practitioners and decision-makers on the ground whose day-to-day work concerns algorithmic, data-driven systems in socially consequential domains.

The tutorial should overview the core tool/framework for the general FAccT audience, situate it in the context where it was evaluated or deployed, analyze the eventual decision about its use, and if applicable, discuss the consequences and ramifications of the tool after deployment.

Suggested List of Topics

Suggested topics for tutorials include but are not limited to the following list.

  • Theoretical, qualitative or quantitative assessment of fairness, discrimination, social equality or vulnerability.
  • Theories of and social and technical mechanisms for individual and collective accountability; Democratic legitimacy and governance mechanisms.
  • Transparency, interpretability, and explainability, e.g. theories of explanation from social sciences and psychology, discussion of necessity, limitations, and real-world uses of existing explanation methods from computer science/Machine Learning literature.
  • Explorations of and introductions to key concepts from Critical Race Theory, Critical Data Studies, Feminist Theory.
  • Methodologies relevant to FAccT issues, including empirical methods in the social sciences; computational methods, e.g., methods and limitations of inferring causality or uncertainty estimation; the contrast between quantitative and qualitative explanations and methodologies; computational methods for collecting and analysing data about socio-technical systems ‘in the wild’, such as web scraping, or citizen science methodologies.
  • Community-based approaches and human factors, e.g. participatory algorithm design; community designed and maintained systems; activism-driven technological change; human-computer interaction; humans-in-the-loop designs; information visualization; user experience design.
  • Perspectives and theory relating to the power exercised through control over data, software, platforms, and protocols; e.g. market concentration, introduction to competing analyses of power.
  • Case studies on the use and impact of automated decision making in industry (e.g. insurance, banking, marketing, media/social media, health, retail) governments, and/or nonprofits.
  • Accounts of the implications of deployment of socio-technical systems, especially in geographies previously under-represented in FAccT, such as Africa, Asia, South America, and Oceania.
  • Explanations and evaluations of regulatory models and their differences, including models originating outside North America and Europe.
  • Guidance for practitioners on how to effectively challenge the development of unethical technologies: e.g. internally, or with tools like strategic litigation or public activism.
  • Anything else of relevance to the FAccT community!

Application Process

The proposal should consist of a maximum of 2 pages + references and must include:

  • The tutorial title. Titles should be in the format, "Translation/Dialogue Tutorial: title" ; “Implications Tutorial: title"; or “Practice tutorial: title”, depending on the type of tutorial being proposed.
  • The tutorial type ("Translation/Dialogue", "Implications" or “Practice”).
  • The team. Given the duration of the tutorials, we encourage tutorials to be presented by one to three people, though more may be involved in its preparation. For the tutorial presenters, please include links to evidence of ability to teach (preferably a video of a past talk or tutorial).
  • A description of the topic you plan to cover, including an impact statement that clearly articulates the main goals and outcomes of the tutorial. Note that this description will be given the highest weight during the selection.
  • A short timeline description of how you plan to break down the material over 45 or 90 minutes. Please also mention here the proposed length (45 or 90 mins), but keep in mind that we might conditionally accept a proposal and suggest a different length. Please explain how you are thinking about the “online experience” aspect of the tutorial.
  • Any important citations
  • A web link or relevant information to the proposed tutorial if it has been delivered previously at any other conference or similar venue.

Submissions must be in PDF format and should ideally be formatted according to the two-column interim ACM Layout Template. Authors who are not familiar with ACM templates may simply submit their tutorial proposals for review in two-column format, with one-inch margins, 9 point Times New Roman font.

Tutorial proposals will be assessed by the Tutorial Chairs, with selections governed by quality and the need for a balanced and diverse program of interest to the FAccT community.

Note (CRAFT): ACM FAccT also solicits proposals on CRAFT (Critiquing and Rethinking Accountability Fairness and Transparency). Some proposals we receive under the CRAFT Call may be more appropriate as proposals for Tutorials, and vice-versa. In such cases, the Tutorial and CRAFT Co-Chairs may transfer such proposals to the other track, in consultation with proposal coordinator(s).