In this reading group, we explore, discuss, and critique data-centric efforts to model large-scale human behavior and belief. Broadly speaking, our focus will be on work that makes advancements in the following areas:
- Application: work that prioritizes using data-centric approaches to understanding, modeling, or predicting specific society-scale dynamics
- Methods: work that introduces new computational, statistical, or machine learning techniques, data sources, and collection tools with clear relevance to modeling society-scale human behavior
- Theory: work that contributes to a deeper, structured understanding of large-scale behavior. Particular attention in this area will be given to work that has some capacity to be operationalized through quantitative methods
We meet every Wednesday afternoon. See below for details!
- Feb 2: Deng et al. Learning Dynamic Context Graphs for Predicting Social Events. KDD, 2019.
- Feb 9: Qiao et al. Learning Evolutionary Stages with Hidden Semi-Markov Model for Predicting Social Unrest Events. Discrete Dynamics in Nature and Society, 2020.
- Feb 16: Halkia et al. Conflict Event Modelling: Research Experiment and Event Data Limitations. AESPEN, 2020.
- Feb 23: et al. The online competition between pro- and anti-vaccination views. Nature, 2020.
Graphical Models and Economics Indicators
- Mar 23: Oberoi et al. A graphical model approach to simulating economic variables over long horizons. Annals of Actuarial Science, 2019.
Social Media Measures of Social Phenomena
- Mar 30: Xi et al. Understanding the Political Ideology of Legislators from Social Media Images. ICWSM 2020.
- Apr 6: Astley et al. Global monitoring of the impact of the COVID-19 pandemic through online surveys sampled from the Facebook user base. PNAS 2021.
- Apr 13: Huberty. Can we vote with our tweet? On the perennial difficulty of election forecasting with social media. International Journal of Forecasting, 2015.
While the topical focus of this reading group is society-scale data science, the content and conduct of our meetings will always reflect a commitment to several central values:
- Respectful discussion: people, ideas, and contributions will be received and treated with respect and consideration.
- Responsible data science: data science applied to society should always be used thoughtfully with the aim of (1) faithfully expressing the underlying human system and (2) improving individual and societal wellbeing - with a genuine and well-informed concern for the intended and unintended impacts work may have.
- Holistic understanding: data science applied to society must be equally committed to understanding the real, human dynamics and drama at play as well as the technical methods available - each from many perspectives. As a result, we place high value on diversity in our reading group, in all its senses including diversity in membership, viewpoints, and disciplinarity of material covered.
- Excellence in communication: ideas live and die by their capacity to engage, inspire, and focus. In both presentations and discussions, we strive individually and collectively to hone our communication skills.
- When does it meet? Wednesdays from 2:45 - 3:45 PM
- Where does it meet? On Zoom until COVID restrictions relax on campus. The weekly Zoom link will be posted in the slack.
- Who can attend? Everyone is welcome. In particular, McGill faculty, staff, graduate students, and undergraduates as well as researchers from the community are all welcome to join weekly meetings.
- How is the reading group coordinated? The reading group will be coordinated through the Slack workspace. Meeting readings, dates, and locations will be shared through the Slack worspace.
Interested? Join the slack! You're also welcome to reach out with any questions to Ben Steel or Prof. Derek Ruths.