Davide Buscaldi

buscaldi [at] lipn.fr
davide.buscaldi [at] polytechnique.edu

I am an Associate Professor (Maître de Conférences, HDR) at Université Sorbonne Paris Nord, in the Knowledge Representation and Natural Language (RCLN) team of the Laboratoire d’Informatique de Paris Nord (LIPN). I also teach as a part-time Lecturer (Professeur Chargé de Cours) at the Computer Science Department of École Polytechnique, and I am a guest member of the DaSciM team at LIX.

My research lies at the crossroads of natural language processing, machine learning and knowledge representation. My current work focuses on large language models (memorization of sensitive information, interpretability, detection of generated text), graph neural networks, and the automatic construction of scientific knowledge graphs, such as CS-KG. Earlier, I worked on geographic information retrieval, word sense disambiguation, question answering, and the detection of irony and humour in text.

I received my PhD in Artificial Intelligence and Pattern Recognition from the Universidad Politécnica de Valencia in 2010, and my HDR from Université Sorbonne Paris Nord in 2025.

selected publications

IJCAI/ECAI

Graph Rewiring in GNNs to Mitigate Over-Squashing and Over-Smoothing: A Survey

Hugo Attali, Nathalie Pernelle, Davide Buscaldi, and 1 more author

In Proceedings of the Thirty-Fifth International Joint Conference on Artificial Intelligence and the 29th European Conference on Artificial Intelligence, IJCAI/ECAI 2026, Bremen, Germany, 15-21 August 2026, 2026

DOI Bib

@inproceedings{DBLP:conf/ijcai/AttaliPBM26,
  author = {Attali, Hugo and Pernelle, Nathalie and Buscaldi, Davide and Malliaros, Fragkiskos D.},
  title = {Graph Rewiring in GNNs to Mitigate Over-Squashing and Over-Smoothing:
                    {A} Survey},
  booktitle = {Proceedings of the Thirty-Fifth International Joint Conference on
                    Artificial Intelligence and the 29th European Conference on Artificial
                    Intelligence, {IJCAI/ECAI} 2026, Bremen, Germany, 15-21 August 2026},
  pages = {7749--7757},
  publisher = {ijcai.org},
  year = {2026},
  url = {https://doi.org/10.24963/ijcai.2026/861},
  doi = {10.24963/IJCAI.2026/861},
  timestamp = {Tue, 22 Sep 2026 16:55:28 +0200},
  biburl = {https://dblp.org/rec/conf/ijcai/AttaliPBM26.bib},
  bibsource = {dblp computer science bibliography, https://dblp.org},
  keywords = {intconf},
}

AAAI

Guess or Recall? Training CNNs to Classify and Localize Memorization in LLMs

J. Dentan, D. Buscaldi, and S. Vanier

In Proceedings of the AAAI Conference on Artificial Intelligence (AAAI 2026), 2026

Oral presentation, Singapore. CORE A*

Bib

@inproceedings{dentan2026guess,
  title = {Guess or Recall? Training CNNs to Classify and Localize Memorization in LLMs},
  author = {Dentan, J. and Buscaldi, D. and Vanier, S.},
  booktitle = {Proceedings of the AAAI Conference on Artificial Intelligence (AAAI 2026)},
  year = {2026},
  note = {Oral presentation, Singapore. CORE A*},
  keywords = {intconf},
}

CVPR

PACT: Pruning and Clustering-Based Token Reduction for Faster Visual Language Models

M. Dhouib, D. Buscaldi, S. Vanier, and 1 more author

In Proceedings of CVPR 2025, 2025

Nashville, USA. CORE A*

Bib HTML

@inproceedings{dhouib2025pact,
  title = {PACT: Pruning and Clustering-Based Token Reduction for Faster Visual Language Models},
  author = {Dhouib, M. and Buscaldi, D. and Vanier, S. and Shabou, A.},
  booktitle = {Proceedings of CVPR 2025},
  year = {2025},
  note = {Nashville, USA. CORE A*},
  keywords = {intconf},
}

ICML

Delaunay Graph: Addressing Over-Squashing and Over-Smoothing Using Delaunay Triangulation

H. Attali, D. Buscaldi, and N. Pernelle

In Proceedings of ICML 2024, 2024

Vienna, Austria. CORE A*

Bib HTML

@inproceedings{attali2024delaunay,
  title = {Delaunay Graph: Addressing Over-Squashing and Over-Smoothing Using Delaunay Triangulation},
  author = {Attali, H. and Buscaldi, D. and Pernelle, N.},
  booktitle = {Proceedings of ICML 2024},
  year = {2024},
  note = {Vienna, Austria. CORE A*},
  keywords = {intconf},
}

ISWC

CS-KG: A Large-Scale Knowledge Graph of Research Entities and Claims in Computer Science

D. Dessì, F. Osborne, D. Reforgiato Recupero, and 2 more authors

In ISWC 2022, LNCS 13489, 2022

pp. 678-696. CORE A

Bib HTML

@inproceedings{dessi2022large,
  title = {CS-KG: A Large-Scale Knowledge Graph of Research Entities and Claims in Computer Science},
  author = {Dessì, D. and Osborne, F. and Reforgiato Recupero, D. and Buscaldi, D. and Motta, E.},
  booktitle = {ISWC 2022, LNCS 13489},
  year = {2022},
  note = {pp. 678-696. CORE A},
  keywords = {intconf},
}

EMNLP

Exploring Vector Spaces for Semantic Relations

K. Gabor, H. Zargayouna, I. Tellier, and 2 more authors

In Proceedings of EMNLP 2017, 2017

Copenhagen, Denmark, pp. 1814-1823. CORE A

Bib HTML

@inproceedings{gabor2017exploring,
  title = {Exploring Vector Spaces for Semantic Relations},
  author = {Gabor, K. and Zargayouna, H. and Tellier, I. and Buscaldi, D. and Charnois, T.},
  booktitle = {Proceedings of EMNLP 2017},
  year = {2017},
  note = {Copenhagen, Denmark, pp. 1814-1823. CORE A},
  keywords = {intconf},
}

SciData

CS-KG 2.0: A Large-scale Knowledge Graph of Computer Science

D. Dessì, F. Osborne, and D. Buscaldi

Scientific Data, 2025

12, 964. Q1 SJR

Bib HTML

@article{dessi2025large,
  title = {CS-KG 2.0: A Large-scale Knowledge Graph of Computer Science},
  author = {Dessì, D. and Osborne, F. and Buscaldi, D.},
  journal = {Scientific Data},
  year = {2025},
  note = {12, 964. Q1 SJR},
  keywords = {journal},
}

FGCS

Generating knowledge graphs by employing Natural Language Processing and Machine Learning techniques within the scholarly domain

D. Dessì, F. Osborne, D. Reforgiato Recupero, and 2 more authors

Future Generation Computer Systems, 2021

116: 253-264. Q1 SJR

Bib HTML

@article{dessi2021generating,
  title = {Generating knowledge graphs by employing Natural Language Processing and Machine Learning techniques within the scholarly domain},
  author = {Dessì, D. and Osborne, F. and Reforgiato Recupero, D. and Buscaldi, D. and Motta, E.},
  journal = {Future Generation Computer Systems},
  year = {2021},
  note = {116: 253-264. Q1 SJR},
  keywords = {journal}
}

DKE

From Humor Recognition to Irony Detection: The Figurative Language of Social Media

A. Reyes, P. Rosso, and D. Buscaldi

Data & Knowledge Engineering, 2012

74: 1-12. Q2 SJR

Bib HTML

@article{reyes2012from,
  title = {From Humor Recognition to Irony Detection: The Figurative Language of Social Media},
  author = {Reyes, A. and Rosso, P. and Buscaldi, D.},
  journal = {Data \& Knowledge Engineering},
  year = {2012},
  note = {74: 1-12. Q2 SJR},
  keywords = {journal}
}

SemEval

SemEval-2018 Task 7: Semantic Relation Extraction and Classification in Scientific Papers

K. Gabor, D. Buscaldi, A.-K. Schumann, and 3 more authors

In Proceedings of SemEval-2018, 2018

New Orleans, USA, pp. 679-688

Bib HTML

@inproceedings{gabor2018semeval,
  title = {SemEval-2018 Task 7: Semantic Relation Extraction and Classification in Scientific Papers},
  author = {Gabor, K. and Buscaldi, D. and Schumann, A.-K. and QasemiZadeh, B. and Zargayouna, H. and Charnois, T.},
  booktitle = {Proceedings of SemEval-2018},
  year = {2018},
  note = {New Orleans, USA, pp. 679-688},
  keywords = {workshop},
}