Marijan
Soric

Ph.D. Candidate

Inria Valda (Inria Paris, DI ENS, CNRS) and Inria CEDAR (Inria Saclay, LIX, CNRS).
École Normale Supérieure  ·  PSL University.

PhD thesis: Exploitation and Structuring of Heterogeneous Geological Data and Knowledge.
Funded by GéolAug (Inria and BRGM, French National Geological Survey).

Supervisors: Pierre Senellart, Ioana Manolescu, Cécile Gracianne.

Previously: École Centrale de Lyon (2021–2024)  ·  NTNU (2024, exchange).

01 Research

Conference Papers

01 KDD 2026  ·  ACM SIGKDD
Benchmarking Table Extraction from Heterogeneous Scientific PDF Documents
Proceedings of the 32nd ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD 2026). ACM, 2026.
Marijan Soric, Cécile Gracianne, Ioana Manolescu, Pierre Senellart.
Paper Code Dataset Website

Preprints

01 arXiv  ·  2025
Benchmarking Table Extraction from Heterogeneous Scientific Documents
Marijan Soric, Cécile Gracianne, Ioana Manolescu, Pierre Senellart.
Preprint Code Dataset Website
02 Master Thesis  ·  2025
Understanding and Extracting Table Information from BRGM Documents
École Centrale de Lyon / Inria, 2025.
Marijan Soric.
HAL Report Slides

Talks

01 Valda Seminar  ·  ENS Paris  ·  23 May 2025
Benchmark: Extracting Table Information from Scientific Documents
Slides Link

Posters

01 KDD 2026 PhD Consortium
Towards Probabilistic Georeferencing of Geological Maps from PDF Reports
Paper
02 BRGM GéoNum Seminar  ·  2025
Benchmarking Table Extraction from Heterogeneous Scientific Documents
Poster
02 Teaching
03 Service
04 Contact

firstname [dot] lastname [at] inria [dot] fr

DI — École Normale Supérieure
45, rue d'Ulm
75 230 Paris Cedex 05, France
Office: Escalier A, Hauts du DI, Bureau 4