Workshop on Critical Data Science
at the 13th International AAAI Conference on Web and Social Media (ICWSM-2019),
Munich, Germany, June 11, 2019
Submissions deadline: March 25, 2019
Acceptance notification: April 12, 2019
————-
The social world is far messier than technical training prepares one for. Among data scientists trained in fields like computer science and statistics are those experiencing a sense of vertigo: we start to realize both the ways in which modeling breaks down on human beings, requiring different notions of rigor, and the potentially negative social impacts of modeling, requiring responsible engagement and activity.
We define “critical data science” as our vision of the practice of working with and modeling data (the “data science”), combined with identifying and questioning the core assumptions commonly underlying that practice (the “critical”). The workshop seeks to combine cultures of critique with those of practice, bringing together data scientists and scholars from computer science and the social sciences around responsibly carrying out data science on social phenomena, and creating sustainable frameworks for interdisciplinary collaboration.
The workshop will involve short reflective presentations by participants, combined with a creative group-based activity to further support reflection of their own and neighboring scientific practices and to create opportunities for further cooperation. The workshop will conclude with a wrap-up for collecting resources and discussing future outcomes, and producing a draft compilation of best practices and a list of priorities for further engagement.
Submissions may either be non-archival 2-page statements of interest or motivation, or archival papers up to 4,000 words. Accepted archival papers will be published in Workshop Proceedings of the 13th International AAAI Conference on Web and Social Media, a special issue of the journal Frontiers in Big Data. Open Access publishing costs will be waived for authors without institutional support for covering these fees.
Topics include:
- What should be standards and practices both of methodological rigor, and of respect for subjects, when carrying out computational research on social systems?
- What role can discussions of methods and instruments play in larger critiques of the limitations of data science?
- What are points of fundamental disagreement or diverging orientations/priorities between disciplines?
- What can we learn from the long tradition of critical scrutiny in statistics?
- What combinations of experiences and/or readings has led data scientists to recognize, and perhaps even adopt, ‘non-technical’ ways of framing the world? How do and can these ways of knowing interact with a modeling approach?
- What philosophical commitments or normative orientations, if adopted by data scientists, would produce a principled data science? How can those be realized in interdisciplinary teams?
- What might it look like to use modeling critically and reflexively, or to contextualize what we can or cannot know from modeling from within the modeling process?
- What can we learn from works looking at the social impact of implemented model-based systems?
- What sorts of practices, coalitions, and collaborations can include marginalized voices into data science rather than exclude them?
- Beyond a space for critical reflection, what can be the positive project of a critical data science?
- How can we design collaborations in critical data science?
See https://projects.iq.harvard.edu/critical-data-science for more information and submission instructions.
Contact: <criticaldatasci2019@gmail.com>.
ORGANIZERS
Momin M. Malik <momin_malik@cyber.harvard.edu>, Berkman Klein Center for Internet & Society at Harvard University
Katja Mayer <katja.mayer@univie.ac.at>, Department of Science and Technology Studies, University of Vienna, and ZSI Centre for Social Innovation Vienna
Hemank Lamba <hlamba@cs.cmu.edu>, School of Computer Science, Carnegie Mellon University
Claudia Müller-Birn <clmb@inf.fu-berlin.de>, Institute of Computer Science, Freie Universität Berlin