June 2026 · Hägerstrand Lab
The Hägerstrand Lab has moved from launch into its first working rhythm. The lab is becoming a shared place for computational human geography: a place where computation is not defined narrowly, but remains open as a family of practices shaped by the people, methods, data, and questions involved.
Looking back at the inauguration
We formally opened the lab on 5 May 2026 in Room 42 at Geocentrum I. The afternoon brought together colleagues and guests for welcoming words, project conversations, fika, and a first shared sense of what the lab can become.
Photos from the afternoon and the invitation are available in the inauguration recap, together with the event invitation.
Master's thesis ideas are welcome
Interested in a master's thesis? If you want to answer established or emerging social-science questions with computational methods, get in touch. We are especially interested in satellite-enabled social science, computational social science, quantitative methods, mobility and register data, and AI-supported text analysis, including qualitative and mixed-methods approaches. We already have a few possible thesis directions and can help shape a topic in the lab's new environment. Write to
philipp.stark@keg.lu.se.
Starting projects and new collaborations
Several research ideas are already emerging around the lab, supported by our close collaboration with the Lund Social Science Methods Centre and its research engineers. This gives the lab a stronger methodological and technical base for projects that need both social-science interpretation and computational infrastructure.
- Large-scale media analysis and policy implementation: We are exploring regional differences in policy implementation through large-scale media analysis, using computational text methods to trace how policy debates and implementation processes vary across places.
- Brazilian Census data and infrastructure news events: Another starting point combines Brazilian Census data with news-event mining around infrastructure projects, connecting structured demographic information with unstructured event narratives.
- Register and cellphone data for knowledge spillovers: A new project idea combines register data with cellphone-based mobility patterns to investigate industrial knowledge spillovers and the spatial dynamics of industrial change.
Talks, papers, and methods in motion
At an AI Lund lunch seminar, we presented a funded project on Tracing Industrial Change Over Time: Harmonizing Industry Classifications in Swedish Register Data for Machine Learning. The work focuses on temporal harmonization of Swedish register data and how changing industrial classifications can be made usable for machine-learning analysis across time.
Our paper on Using Text-Based Life Trajectories from Swedish Register Data to Predict Residential Mobility with Pretrained Transformers has been published in the proceedings of PAKDD 2026 and was presented at the conference in Hong Kong. The work transforms longitudinal register records into text-based life trajectories, making it possible to use modern NLP and transformer models to study long-term life-course dynamics at population scale.
C2S Nordics and the wider community
We are preparing to take part in C2S Nordics. Abstracts have been submitted, and we hope to use the event to connect the Hägerstrand Lab with the wider Nordic computational social science community.
Biweekly meetings
The lab's biweekly meetings have started. For now, they are a practical space for exchange, coordination, and learning from ongoing work. In the future, we plan to extend the format by inviting more people and hosting invited speakers around computational methods, data infrastructures, and social-science applications.
Back in August
After the summer break, we will return in August with more details on the first projects, thesis opportunities, meetings, and collaborations. Until then, we wish everyone a good summer and a happy midsummer.