CMSC 25900: Ethics, Fairness, Responsibility, and Privacy in Data Science

Week	Date	Topics	Instructor	Due Dates	Readings and Assignments
1	4/7	Course Overview and Introduction to the Process of Data Science	Blase and Raul	--	Reading 1: Solon Barocas and Andrew D. Selbst. Big Data's Disparate Impact. 104 California Law Review 671, 2016. Reading 2: Rob Schmitz. In Germany, High Hopes For New COVID-19 Contact Tracing App That Protects Privacy. NPR, April 2, 2020. Reading 3: Lorenzo Franceschi-Bicchierai. We Saw NSO's Covid-19 Software in Action, and Privacy Experts Are Worried. Vice Motherboard, April 2, 2020.
1	4/9	Data Acquisition and Standardization (descriptive statistics, assumptions about data formats, dirty data, missing data)	Raul	Reading Response 1 (4/9)
2	4/14	Data Analysis (inferential statistics, sampling, confidence intervals, hypothesis testing, pitfalls)	Raul	Reading Response 2 (4/13), Assignment 1 (4/15)	Reading 1: danah boyd and Kate Crawford. Critical Questions for Big Data. Information, Communication, and Society. 2012 Reading 2: Stephen Stigler. Data Have a Limited Shelf Life. HDSR 2019. Reading 3: Michael Jordan Artificial Intelligence: The Revolution Hasn't Happened Yet. HDSR 2019. Programming Assignment 1 introduces basic data exploration, analysis, and recoding in Python.
2	4/16	Introduction to Machine Learning and Generalization (linear/multiple regression, goodness of fit, extrapolation, decisions)	Raul	--
3	4/21	Additional Methods for Machine Learning and Feature Engineering (feature engineering/model selection, classifiers, pitfalls)	Raul	Reading Response 3 (4/20), Assignment 2 (4/22)	Reading 1: Julia Angwin, Jeff Larson, Surya Mattu, Lauren Kirchner. Machine Bias. ProPublica, May 23, 2016. Reading 2: Solon Barocas, Moritz Hardt, Arvind Narayanan. Fairness and Machine Learning Chapter 2: Classification. fairmlbook.org, 2019. Reading 3: Solon Barocas, Moritz Hardt, Arvind Narayanan. Fairness and Machine Learning Chapter 5: Testing Discrimination in Practice. fairmlbook.org, 2019. Programming Assignment 2 introduces the difficulties of cleaning, joining, and generalizing data.
3	4/23	Fairness in Machine Learning and Artificial Intelligence	Blase	--
4	4/28	Experimental Design in Data Science (hypothesis testing, experimental design, RQ, multiple comparisons)	Blase	Reading Response 4 (4/27), Assignment 3 (4/29)	Reading 1: Roy A. Maxion. Making Experiments Dependable. In Dependable and Historic Computing, Lecture Notes in Computer Science Volume 6875, pp. 344-357, 2011. Reading 2: Arunesh Mathur, Gunes Acar, Michael J. Friedman, Elena Lucherini, Jonathan Mayer, Marshini Chetty, Arvind Narayanan. Dark Patterns at Scale: Findings from a Crawl of 11K Shopping Websites. In Proc. CSCW, 2019. Reading 3: Jamie Luguri and Lior Strahilevitz. Shining a Light on Dark Patterns. University of Chicago, Public Law Working Paper No. 719, 2019. Programming Assignment 3 covers fairness issues in applied machine learning.
4	4/30	The Design of Respectful User Interfaces for Data (dark patterns, nudging, manipulative interfaces, respectful interfaces)	Blase	--
5	5/5	Responsible and Ethical Data Collection (ethics of human subjects, data subjects, sampling/recruitment, scraping)	Blase	Reading Response 5 (5/4), Assignment 4 (5/6)	Reading 1: Department of Health, Education, and Welfare. The Belmont Report. April 18, 1979. Reading 2: Michelle N. Meyer. Everything You Need to Know About Facebook's Controversial Emotion Experiment. Wired, June 30, 2014. Reading 3: Robinson Meyer. Everything We Know About Facebook's Secret Mood Manipulation Experiment. The Atlantic, June 28, 2014. Programming Assignment 4 requires students create an interface exhibiting a dark pattern and compare its efficacy relative to a control interface in an online experiment.
5	5/7	Responsibly Visualizing and Communicating Data	Raul	--
6	5/12	Privacy Philosophy, Regulation, and Policy	Blase	Reading Response 6 (5/11), Assignment 5 (5/13)	Reading 1: Cynthia Dwork and Aaron Roth. The Algorithmic Foundations of Differential Privacy (only pages 5-33), 2014. Reading 2: Jeffrey Mervis. Can a Set of Equations Keep U.S. Census Data Private? Science Magazine, January 4, 2019. Reading 3: Daniel Solove. 'I've Got Nothing to Hide' and Other Misunderstandings of Privacy. San Diego Law Review 44, 2007. Programming Assignment 5 explores how to visualize data in responsible and irresponsible ways.
6	5/14	Statistical Approaches to Privacy (k-anonymity, differential privacy)	Raul	--
7	5/19	Anonymization and Anonymity	Blase	Reading Response 7 (5/18), Assignment 6 (5/20)	Reading 1: Paul Ohm. Broken Promises of Privacy: Responding to the Surprising Failure of Anonymization. UCLA Law Review Vol. 57, p. 1701, 2010. Reading 2: Arvind Narayanan and Vitaly Shmatikov. Robust De-anonymization of Large Sparse Datasets. In Proc. IEEE Symposium on Security and Privacy, 2008. Programming Assignment 6 explores the implementation and limits of differential privacy.
7	5/21	Conducting and Explaining Data Analysis Using Neural Networks	Blase	--
8	5/26	Responsible Data Lifecycles (the right of erasure, access rights, data portability, biomedical and genetic data, web personalization)	Blase	Reading Response 8 (5/25), Assignment 7 (5/27)	Reading 1: Theo Bertram et al. Five Years of the Right to be Forgotten. In Proc. CCS, 2019. Reading 2: Rebecca Skloot. 'Henrietta Lacks': A Donor's Immortal Legacy. NPR, February 2, 2010. Reading 3: Megan Molteni. The Creepy Genetics Behind the Golden State Killer Case. Wired, April 27, 2018. Programming Assignment 7 provides hands-on experience scraping data from the web and then using that information to deanonymize a data set.
8	5/28	Emerging Uses and Misuses of Data (data brokers, data sharing, future uses)	Raul
9	6/2	Should You Ask This Question? Should You Collect This Data?	Raul	Project Report (6/1), Reading Response 9 (6/1), Assignment 8 (6/3)	No new readings, though there still is a course-synthesis-style reading response. Programming Assignment 8 requires students to simulate a database-backed application that collects data robustly while also complying with GDPR-style rights to data access and erasure.

Schedule and Assignments (Spring 2020)