Data Ethics
From issues of disinformation to algorithmic bias to inequalities in data privacy violations, we find ourselves today increasingly confronting pressing questions about the ethics of information and data. Focused both on broader social issues and applications in particular domains of data-driven research, the goal of the University of Oregon Data Ethics Thrust is to both raise awareness about and build usable tools for addressing manifold questions of data ethics today. This page offers information about some of our ongoing projects and resources.
Responsible Data Science – Workshop Series
-
- #1: Algorithmic Bias Workshop
- #2: Equality in Data Curation Workshop
- #3: Data Privacy Workshop
Data Ethics – Conversation Series
Responsible Data Science Workshop Series
Are you a faculty, postdoctoral, or graduate student researcher making use of data science methods in your work? Are you concerned about, or maybe even just curious about, algorithmic fairness and possible machine bias influences on your uses of data sets? If so, we have developed a workshop specifically targeted for you. The “Responsible Data Science” series, running Winter and Spring 2022, is a joint venture of the UO Data Science Initiative, Department of Philosophy, School of Law, and Oregon Health Sciences University.
Algorithmic Bias Workshop
Our first workshop was held on Friday, February 18th at 2:00 in LLCS (Living Learning Center South) 101, on the topic of Algorithmic Bias. This workshop offers a set of conceptual tools for identifying forms of bias at different stages of the machine learning pipeline. We will also explore strategies for mitigating the unwanted effects of algorithmic bias. If these issues are relevant to your work, please come, bring a computer, and be prepared with questions.
The workshop was developed and conducted by Ramón Alvarado (UO Philosophy) and Paul Showler (UO Philosophy); a re-recording of the workshop’s content (produced by Dr. Alvarado) is available directly below.
Equality in Data Curation Workshop
Workshop #2 was held on Friday, April 22nd at 2:00p in Knight Library Room 101. The focus was on tools for understanding the conceptual underpinnings and historical contexts of data structures, fields, and variables defining our databases. There was discussion of the value of equality and the very meaning of equality itself (does equality imply sameness, difference-sensitive treatment, or something else?). Also presented were tools for assessing and identifying potential inequalities in data design itself.
The workshop was developed and conducted by Colin Koopman (UO Philosophy) and Steven Bedrick (OHSU Medical Informatics and Clinical Epidemiology) with assistance from Paul Showler, Brooke Burns, and Asher Caplan (all of UO Philosophy). A recording of the first portion of the workshop (focusing on Dr. Koopman’s slides) is available directly below (recommended viewing is via the ‘Panopto’ viewer by clicking on the up arrow near the bottom of the player, as this view will enable you to adjust the relative sizes of a dual-screen slideshow plus video recording).
Data Privacy Workshop
Workshop #3 was held on Friday, May 20th at 3:00p in Hedco Education Bldg Room 220. Presenters covered four topics in privacy ethics: 1) the meanings and understandings of privacy, 2) particular toolsets, or analytics, for interrogating potential privacy biases in datasets, and 3) hands-on application of these analytics to real-world (and maybe also a few fictionalized) cases, and 4) more general societal challenges facig privacy. The analytic presented in part 2 is based on the taxonomy of privacy developed by Deirdre Mulligan, Colin Koopman, and Nick Doty in their 2016 article in Philosophical Transactions of the Royal Society A (link). Case studies used in the workshop are available here (case studies document).
This workshop was developed by Ramón Alvarado (UO Philosophy) and Colin Koopman (UO Philosophy), with the assistance of Philosophy Ph.D. students Paul Showler and Brooke Burns.
Additional future workshop topics include: Research Ethics, Database Interoperability Ethics, Deeper Details on Data Privacy, & more (email us a request)
Data Ethics Conversation Series
The Philosophy Department at the University of Oregon is pleased to host its second webinar conversation series on Data Ethics. Data-driven methodologies, practices and technologies are ubiquitous in scientific inquiry and policy-making contexts. The socio-technical implications and limitations of their use are now gradually being acknowledged as deep and complex in a lively . This webinar series on data ethics will provide an overview of the fundamental philosophical, ethical and pragmatic issues related to the practice of data science and associated computational methodology (machine learning, artificial intelligence, etc.).
For any questions regarding scheduling and registration please visit ramonalvarado.net or contact Ramón Alvarado at ralvarad@uoregon.edu with subject line “data ethics webinar.”
Spring 2022 Program 
Find presentation abstracts here.
April
Monday, April 18th 10:00am
Topic: AI Ethics and Investors
Speaker: Ravit Dotan (UC-Berkely)
May
Wednesday, May 4th 10:00am
Topic: Target specification bias in healthcare applications of machine learning
Speaker: Eran Tal (McGill)
Wednesday, May 18th 10:00am
Topic: Epistemic Risk and Modeling Decisions in Science
Speakers: Stephanie Harvard (UBC) and Eric Winsberg (University of South Florida)
June
Wednesday, June 1st 10:00am
Topic: Using (un)fair algorithms in an unjust world
Speaker: Kasper Lippert-Rasmussen
(University of Aarhus/Arctic University of Norway)
Monday, June 6th 10:00am
Topic: Automated Decision-Making as Domination:
what’s missing from discussions of AI & justice?
Speaker: Jenna Burrell (Data and Society)