| _version_ | 1866902166802792448 |
|---|---|
| author | Bernard, Gaël |
| author_facet | Bernard, Gaël |
| contents | <p><em><span lang="EN-GB">Slides of a Lightning Talk held at the first Swiss Open Academic Data (SOAD) Day, 10 September 2025.</span></em></p> <p><span lang="EN-GB">Abstract: At EPFL, we developed a classifier to assess whether scientific publications are relevant to particular research topics such as Horizon Europe calls. Initially applied to EPFL’s internal repository, the tool helps identify which labs are most aligned with current funding opportunities. This talk will provide a high-level overview of how the classifier works — combining embedding-based filtering with ChatGPT-based relevance scoring — and discuss how it can be extended to open data sources, such as OpenAlex, to support broader discovery and benchmarking. A key focus will be on the complementary use of internal institutional data (where the structure and lab affiliations are well known) and external open sources like OpenAlex (which enable comparisons across institutions). We’ll reflect on what institutions can gain from leveraging both perspectives, from improved grant targeting to stronger strategic insights.</span></p> |
| format | Recurso digital |
| id | zenodo_https___doi_org_10_5281_zenodo_17255145 |
| institution | Zenodo |
| language | |
| publishDate | 2025 |
| publisher | Zenodo |
| record_format | zenodo |
| spellingShingle | Using AI to Align Research with Grants: From Institutional Data to OpenAlex Bernard, Gaël <p><em><span lang="EN-GB">Slides of a Lightning Talk held at the first Swiss Open Academic Data (SOAD) Day, 10 September 2025.</span></em></p> <p><span lang="EN-GB">Abstract: At EPFL, we developed a classifier to assess whether scientific publications are relevant to particular research topics such as Horizon Europe calls. Initially applied to EPFL’s internal repository, the tool helps identify which labs are most aligned with current funding opportunities. This talk will provide a high-level overview of how the classifier works — combining embedding-based filtering with ChatGPT-based relevance scoring — and discuss how it can be extended to open data sources, such as OpenAlex, to support broader discovery and benchmarking. A key focus will be on the complementary use of internal institutional data (where the structure and lab affiliations are well known) and external open sources like OpenAlex (which enable comparisons across institutions). We’ll reflect on what institutions can gain from leveraging both perspectives, from improved grant targeting to stronger strategic insights.</span></p> |
| title | Using AI to Align Research with Grants: From Institutional Data to OpenAlex |
| url | https://doi.org/10.5281/zenodo.17255145 |