EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Heterogeneous Graph-Based Partner Recommendation for EU Framework Programmes

Jeovane Honorio Alves, Stefan Morcov, Radu State

IEEE Access · 2026 · 14 · 94296-94313

DOI: 10.1109/access.2026.3702043 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Access, 2026. Verfasst von Jeovane Honorio Alves und 2 weiteren Autor:innen. Die Arbeit nennt 39 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Luxembourg National Research Fund.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Assembling competitive consortia for EU Framework Programme proposals requires identi- fying partners that are thematically relevant, institutionally compatible, and programmatically aligned — yet partner search typically defaults to existing networks, narrowing the search space. We formulate seeded partner recommendation — conditioned on a target project and one known consortium member — as a ranking problem over a heterogeneous knowledge graph constructed from CORDIS data (FP1 through Horizon Europe), comprising around 504K nodes and 1.55M typed edges. The model integrates collaboration history, thematic topics, legal and funding context, geography, and text embeddings, and is evaluated under time-aware splits simulating forward-looking recommendation. Among seven GNN operators, SAGEConv achieves MRR of 35.04% and Hits@10 of 62.43% in warm-start, and MRR of 33.56% and Hits@10 of 61.39% in cold-start. Ablation studies show that legal basis, funding scheme, and project-side semantics are the dominant signals for recommendation quality.

Förderung

  • Luxembourg National Research Fund

Zitieren

APA

Alves, J. H., Morcov, S., & State, R. (2026). Heterogeneous Graph-Based Partner Recommendation for EU Framework Programmes. IEEE Access, 14, 94296-94313. https://doi.org/10.1109/access.2026.3702043

Vancouver

Alves JH, Morcov S, State R. Heterogeneous Graph-Based Partner Recommendation for EU Framework Programmes. IEEE Access. 2026;14:94296-94313. doi:10.1109/access.2026.3702043

BibTeX
@article{Alves2026,
  author = {Alves, Jeovane Honorio and Morcov, Stefan and State, Radu},
  title = {Heterogeneous Graph-Based Partner Recommendation for EU Framework Programmes},
  journal = {IEEE Access},
  year = {2026},
  volume = {14},
  pages = {94296--94313},
  doi = {10.1109/access.2026.3702043}
}
Kostenlos starten →