Social Media Analysis Methods and Applications

Course Description   Assignments   Schedule

• Instructor: Silvio Amir
• Meetings: Tue & Fri 9:50am - 11:30am @ Ryder Hall 153
• Office hours: Thur 12pm - 2pm @ Office 2207, 177 Huntington Avenue and @ Zoom
• Discussion/Announcements: piazza

Course Description

Seminar class in which we read, critique and discuss recent research papers about the methods, applications and implications of social media analysis. Social media analysis is a multidisciplinary subject that encompasses a broad range of topics. In this class we will focus on: (i) cutting-edge Machine Learning/Deep Learning and Natural Language Processing methods to infer signals from single posts, individual users, and user networks; (ii) applications of social media analysis to real-world analyses in domains such as the social sciences, political science, mental-health and epidemiology; and (iii) the implications, as well the ethical and moral considerations, of deploying social media analysis systems and Human-centered AI systems more broadly.

Students will take turns presenting a paper each week (2 presentations per class) which we will discuss together critically during class meetings. The presentations should also cover any relevant background material; this is important because in general we will have a strong bias toward very recent work. Critical discussion will follow, and we will conclude class by talking about potential means of extending the work. Students will also be responsible for writing short summaries of the research papers we read each week. Student led presentations will be complemented with several guest lectures from top researchers working on different facets of social media analysis.

In addition to the presentations and paper reviews, students will propose and execute a mini project culminating in a research paper (draft). The results and findings of the project will be presented in a class, structured as a research talk. Therefore, this class will very much be research-oriented.

⚠️ As per university policy, the class the will be in-person only and everyone must be wearing a mask.

Prerequisites

Students should have an interest in conducting (or learning how to conduct) research. There are no official pre-reqs for this class, however students are expected to have some background in data science, machine learning and statistical natural language processing. This is mainly to make students life easier since we will not cover background knowledge related to the papers in class.

Assignments

Oral Presentations

Students will rotate responsibility for presenting (one of the) assigned papers and leading the ensuing discussion. The presentation should provide necessary background, the core contribution of the paper, (perceived) strengths and weaknesses, and ideas for improving the work.

Written Critiques

Prior to every class all students are to write a brief “review” of one of the assigned papers for that class (their choice). These should include a concise summary of the work and enumerate at least two strong points and at least two weak points therein. Furthermore, these critiques will note ways in which the work could potentially be extended/improved (which may be coupled with the weak points). Assessments are to be submitted via canvas by 11.59pm of the day prior to class (i.e. Mondays and Thursdays). Note that is assumed students read all the assigned papers, but may read only one in-depth. The idea is basically to ensure robust discussion during class meetings.

Project

The class will culminate in projects with accompanying papers (along with the respective code, when possible). The papers will be structured as research papers following a template, available here and the code should be submitted as a jupyter notebook (if possible). Students may work together but then a concomitant increase in complexity and a clear delineation of contributions is expected, for fairness sake.

Students will be able to choose which projects to work on. Projects may involve replicating a state-of-the-art paper we read, or, ideally, extending one of these or developing new ideas in the area. Students are to write a short (1 page) project proposal including the motivation, goals, expected results and a timeline of key milestones. Project proposals will then be reviewed and approved by the instructor. Some class time will be reserved for students to present their project ideas, results and findings (these will be structured as research talks).

Projects will be evaluated for their relevance, scientific rigour, and originality. The papers will be evaluated as paper draft submissions using the template available here.

Grades

• 55% Project and write-up
• 20% In class presentations of papers
• 15% Participation
• 10% Written summaries/critiques of papers (on canvas)

Schedule

Tentative weekly schedule. The assigned papers are unlikely to change but the presentation dates might move to accommodate the (busy) schedules of the invited speakers.