About

I am a PhD Candidate in Artificial Intelligence at Institut Polytechnique de Paris (Télécom Paris), under the supervision of Thomas Bonald (LTCI – Information Processing and Communications Laboratory, Télécom Paris) and Tiphaine Viard, (i3 – Interdisciplinary Institute of Innovation, Télécom Paris).
In my research, I study enriched graph methods and algorithms to leverage information from real-world data, in order to add transparency in current machine learning approaches.

My research interests include:

  • Real-world graphs
  • Graph algorithms
  • Graph Neural Networks
  • Network analysis in the social sciences
  • Fairness and ethics in AI

Research

PUBLICATIONS

Evaluating Graph-Based Explanations for AI-Based Recommender Systems
Simon Delarue, Astrid Bertrand, Tiphaine Viard
[Under review]

Graph Neural Networks: The Surprising Superiority of Logistic Regression
Simon Delarue, Thomas Bonald, Tiphaine Viard
[Under review]

Mapping AI Ethics: a Meso-Scale Analysis of Charters and Manifestos
Mélanie Gornet, Simon Delarue, Maria Boritchev, Tiphaine Viard
ACM Conference on Fairness, Accountability and Transparency, 2024

Interpretive Flexibility in Data Science and Artificial Intelligence
Tiphaine Viard, Mélanie Gornet, Simon Delarue
Undone Computer Science Conference, 2024
[paper]

Sparse Graph Neural Networks with Scikit-network
Simon Delarue, Thomas Bonald
International Conference on Complex Networks and their Applications, 2023
[code] [paper]

Unexpected Attributed Subgraphs: a Mining Algorithm
Simon Delarue, Tiphaine Viard, Jean-Louis Dessalles
IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining, 2023
[code] [paper]

TALKS

HI!PARIS READING GROUP ON GRAPH NEURAL NETWORKS, 2022
Representation Learning on large graphs (with Thomas Bonald)
[ressource material]

INSTITUT POLYTECHNIQUE DE PARIS DOCTORAL SCHOOL DAY, 2022
Unexpected attributed subgraphs: a mining algorithm (Best poster award in the field of Computing, Data and Artificial Intelligence)
[poster]

HI!PARIS SUMMER SCHOOL, 2022
Sparse Graph Neural Networks (Best poster award)
[poster]

AWARDS

  • Best Poster Award Unexpected attributed subgraphs: a mining algorithm
    [poster]
  • Best Poster Award Sparse Graph Neural Networks
    [poster]