Scientific knowledge representation
How far can social science knowledge be captured in machine-readable knowledge graphs, and what can be extracted from publications reliably and at scale?
Computational Social Science
Simon Walo
SNSF Ambizione Fellow · Social Networks Lab, ETH Zürich
I build knowledge graphs of science. Using large language models, I turn scientific texts into structured knowledge to study how innovation emerges, and to imagine better ways of sharing what we know.
Research agenda
More about my researchScience of science
How do innovations emerge, diffuse and shape the development of knowledge? Knowledge graphs let us study the content of research, not just its metadata.
The future of scientific publishing
What would a publication system built on machine-readable knowledge look like, and how can the transition be organized successfully?
Current project · SNSF Ambizione 2026–2030
Causal Networks and Innovation Patterns in the Social Sciences
Scientific innovation is usually measured as a single dimension, without knowing what kind of innovation it is or where it sits in existing knowledge. This project extracts causal claims from millions of social science abstracts with open-source language models and assembles them into a large causal network. This makes it possible to tell apart new causes, new effects, new links and new concepts, and to study how their place in the network shapes their impact.
DataOpenAlex, Web of Science
MethodLLM-based information extraction and network analysis
HostSocial Networks Lab, ETH Zürich
Selected publications
All publications-
2023
‘Bullshit’ After All? Why People Consider Their Jobs Socially Useless
-
2023
The link between routine tasks and job polarization: A task measurement problem?DOI 🏆 Top Viewed Article · Wiley (2025)
-
2024
Understandings of Culture: A European Bottom-Up Study Using Structural Topic Modelling