This course focuses on the cybersecurity threats that emerge when humans interact with technology at scale – such as misinformation, disinformation, and online harassment. We will study how these threats emerge in practice and explore the landscape of technical defenses and interventions against these harms. Students will primarily read, synthesize, present, and discuss research papers. The course will culminate in a presentation of a quarter-long research project that students will conduct in small groups.
Lecture: Tu/Th 12:30–1:50 PM. Warren Lecture Hall 2207.
Instructor: Deepak Kumar
Office Hours: Tuesday 2:00–3:00 PM, or by appointment, CSE 3248.
Teaching Assistant: Muchan Li
Office Hours: By appointment.
Prerequisites: Some cybersecurity knowledge and human-computer interaction knowledge is helpful, but not required. This is a research focused course, so students who are excited to explore novel topics are encouraged to enroll.
Communication: We use Canvas for announcements; discussion happens on Piazza. You can find the syllabus here.
Submissions: All course assignments will be submitted through Gradescope. Entry Code: NY4WBJ.
The tentative schedule and readings for the class are below:
[DK] Course introduction, definitions, harms, and a brief history. [slides]
[DK] The harm landscape, the structure of Trust & Safety inside companies. [slides]
[DK] How do we measure online safety in research and in practice? What metrics do we use to communicate success? How do we do research in this space? [slides]
Commericial Content Moderation, by Sarah Roberts. From her book, "Behind the Screen: Content Moderation in the Shadows of Social Media" [link]
Metaphors in Moderation, by Joseph Seering, Geoff Kaufman, and Stevie Chancellor. In New Media & Society, 2020. [link]
Human-AI Collaboration via Conditional Delegation: A Case Study of Content Moderation, by Vivian Lai et al. CHI 2022. [link]
Comparing the Perceived Legitimacy of Content Moderation Processes: Contractors, Algorithms, Expert Panels, and Digital Juries, by Christina Pan et al. CSCW 2022. [link]
Primer, everyone should read. Disinformation’s spread: bots, trolls and all of us, by Kate Starbird. Nature World View. [link]
Like-minded sources on Facebook are prevalent but not polarizing, by Brendan Nyhan et al. Nature 2023. [link]
On the infrastructure providers that support misinformation websites, by Catherine Han et al. AAAI ICWSM 2022. [link]
Specious Sites: Tracking the Spread and Sway of Spurious News Stories at Scale, by Hans Hanley et al. IEEE Security & Privacy 2024. [link]
Tactics, Threats, and Targets: Modeling Disinformation and its Mitigation, by Shujaat Mirza et al. ISOC Network and Distributed Systems Security Symposium 2023. [link]
SoK: Hate, Harassment, and the Changing Landscape of Online Abuse, by Kurt Thomas et al. IEEE Security & Privacy 2021. [link]
Designing Toxic Content Classification for a Diversity of Perspectives, by Deepak Kumar et al. USENIX Symposium on Usable Privacy and Security. [link]
"You Know What to Do”: Proactive Detection of YouTube Videos Targeted by Coordinated Hate Attacks, by Nerico Mariconti et al. ACM CSCW 2019. [link]
Jury Learning: Integrating Dissenting Voices into Machine Learning Models, by Mitchell Gordon et al. ACM CHI 2022. [link]
Understanding the Digital Lives of Youth: Analyzing Media Shared within Safe Versus Unsafe Private Conversations on Instagram, by Shiza Ali et al. ACM CHI 2022. [link]
Mental Models, Expectations and Implications of Client-Side Scanning: An Interview Study with Experts, by Divyanshu Bhardwaj et al. ACM CHI 2024. [link]
Profiling the Offline and Online Risk Experiences of Youth to Develop Targeted Interventions for Online Safety, by Ashwaq Alsoubai, et al. ACM CSCW 2024. [link]
Experimental Analyses of the Physical Surveillance Risks in Client-Side Content Scanning, et al. ISOC NDSS 2024. [link]
The spyware used in intimate partner violence, by Rahul Chatterjee et al. IEEE Security & Privacy 2019. [link]
Clinical Computer Security for Victims of Intimate Partner Violence, by Sam Havron et al. USENIX Security 2019. [link]
The Tools and Tactics Used in Intimate Partner Surveillance: An Analysis of Online Infidelity Forums, by Emily Tseng et al. USENIX Security 2020. [link]
The Digital-Safety Risks of Financial Technologies for Survivors of Intimate Partner Violence, by Rosanna Bellini et al. USENIX Security 2023. [link]
Reconsidering tweets: Intervening during tweet creation decreases offensive content, by Matt Katsaros et al. AAAI ICWSM 2022. [link]
You can't stay here: The efficacy of reddit's 2015 ban examined through hate speech, by Eshwar Chandrasekharan et al. ACM CSCW 2017. [link]
[DK] Lecture on a topic selected by the class via this form: https://forms.gle/Bjj4zthaw9Ghbhqm6
Presentations on the quarter-long project.
Schedule is here: https://docs.google.com/spreadsheets/d/1GrM84jblQIZ3Tkq_Mv5YXCojzLtQrGwOaS-VLV_9fJU/edit?gid=0#gid=0
Presentations on the quarter-long project.
Schedule is here: https://docs.google.com/spreadsheets/d/1GrM84jblQIZ3Tkq_Mv5YXCojzLtQrGwOaS-VLV_9fJU/edit?gid=0#gid=0
This is a discussion oriented research course. Students will be responsible for reading and discussing papers in class, presenting at least one paper during the quarter, and working on a quarter-long project in groups of 2-3. Grading is based on the following components:
Attendance and Participation (25%)
Attendance is mandatory in class. The expectation is that all students will have done the reading in advance of class. To kick off the class and get everyone on the same page, Deepak will begin each paper-discussion session with cold calling students at random with a pre-selected set of questions. These questions are not necessarily to test comprehension, but rather, to get students thinking critically about the reading material and provide the necessary background to begin discussing the paper.
Paper Presentation and Discussion Lead (25%) [Presentation Details], [Paper Assignments]
Students will be assigned a paper to lead discussion on with a set of slides by week 2 of the quarter. Students will be expected to share the slides with Deepak at least 3 days in advance and iterate with him on developing the final slides before the presentation date. More details about the structure of this assignment will follow in lecture.
Term Project (50%) [Project Details]
Students will work on an independent research project over the term in groups of 2-3, which will culminate in a project presentation and small (~5 page) writeup. The grade is divided into several subparts: