Information Extraction of Doctoral Theses Using Two Different Large Language Models vs Health Services Researchers: Development and Usability Study
Jonas Cittadino, Pia Traulsen, Teresa Schmahl, Larisa Wewetzer, Julia Cummerow, Kristina Flägel, Christoph Strumann, Katja Goetz, Jost Steinhäuser · 2025
Read Information Extraction of Doctoral Theses Using Two Different Large Language Models vs Health Services Researchers: Development and Usability Study by Jonas Cittadino, Pia Traulsen, Teresa Schmahl, Larisa Wewetzer, Julia Cummerow, Kristina Flägel, Christoph Strumann, Katja Goetz, Jost Steinhäuser (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-12694942.txt.
Description
Jonas Cittadino and co-authors test whether large language models can pull key facts from old doctoral theses in a 2025 study. An archive in Germany holds about five hundred paper theses in family medicine, some dating back to 1965. The team picked ten theses at random from across the years. Two model systems, one from OpenAI and one from Google, were set to extract facts and draft summaries.
Both systems found the core facts, such as author, place, and year, for all ten theses. One system wrote summaries for all ten, and the other for nine. Blind raters scored the machine summaries much like the human one. Making them was many times faster than human work. The authors note the risk of made-up facts as a caveat. They still see the models as a real help for archives.
Catalog record
| CREATOR | Jonas Cittadino, Pia Traulsen, Teresa Schmahl, Larisa Wewetzer, Julia Cummerow, Kristina Flägel, Christoph Strumann, Katja Goetz, Jost Steinhäuser |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 28 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |