The App Team at the LITIS Lab boasts significant expertise in graph-based machine learning, a field we have been engaged in for many years. Our team’s focus has intensified recently, especially due to the burgeoning interest in machine learning applications for graph data. Collaborative efforts with the GREYC and LIFAT labs have further helped us consolidate our focus in this domain.

Our involvement in the AGAC project marked a pivotal point, leading to notable discoveries in the realm of Graph Neural Networks, as documented in our publications [1, 2]. Building on these achievements, we embarked on the ambitious ANR CodeGNN project. This project aims to delve into three key areas: convolution techniques, pooling methods in Graph Neural Networks (GNNs), and the application of GNNs to spatio-temporal graph data.

Over the past year, we have fostered a dynamic environment of collaboration and knowledge exchange through a working group. This group regularly discusses progress and insights not only from our consortium’s teams but also from other leading teams at the intersection of graph theory and machine learning. For more information about our meetings and discussions, please visit the NormaSTIC website here.

We are always eager to engage with fellow researchers. If you’re reading this and are interested in sharing your work, we warmly invite you to reach out. We would be delighted to arrange an online seminar or welcome you to Rouen for a presentation.