LabFam Research Center Workshop by Prof. Ingmar Weber (Saarland University)

On behalf of the LabFam Research Center, we cordially invite you to attend a workshop led by Prof. Ingmar Weber (Saarland University). The researcher holds an Alexander von Humboldt Professorship for AI, Germany’s most prestigious research award. He is also the founder and co-director of the Interdisciplinary Institute for Societal Computing (I2SC). In his research, he explores how novel data sources - ranging from ChatGPT logs to satellite imagery - can be used to study societal phenomena.

The workshop will take place at the Faculty of Economic Sciences on Thursday, 2 April, in Room A409. It will begin at 09:00 and conclude at 16:15, including coffee and lunch breaks. A brief description of the workshop, along with the schedule, is provided below.

[The workshop will be held in English.]

---
Workshop abstract: This one-day workshop explores how researchers can use novel data sources, from advertising audience estimates to satellite imagery, to study timely and policy-relevant societal phenomena. Across four 90-minute sessions, we will go through the full pipeline: data access and ethics, bias and validation, and how to turn a research question into a feasible empirical design. Case studies will mostly focus on topics related to gender and mobility, but the emphasis will be on the data sources and the analysis methods.

Workshop structure:

Session 1: Everything Is Data: How to Tap Into New Data Sources for Computational Social Science
This session provides the motivation for and introduction to the use of unconventional data to study society. It also covers the essential methodological and ethical considerations of collecting “open” data.

Session 2: Gender Gaps and Relationships in the Digital Age
A deep dive into case studies of digital gender gaps, relationship breakups, and AI-human relationships. Data sources covered include advertising audience estimates, prices on Cameo, bios on Twitter, multi-year ChatGPT logs, and posts on the subreddit r/replika.

Session 3: Counting People on the Move from a Distance
A deep dive into case studies of international migration, internal displacement, and point-of-interest busyness. Data sources covered include advertising audience estimates, very-high-resolution satellite images, medium-resolution satellite images, geotagged tweets, and data from Google Maps.

Session 4: Idea-to-Study Clinic: Feasibility, Data, Identification, and Ethics
Attendees are invited to share a short idea at least one week before the event. During this session, Ingmar will provide feedback and feasibility analyses concerning the use of novel data sources based on these submissions.