Hi I’m Manuel Dileo and I’m a Postdoctoral Researcher at Human Technopole, working in the Center for Computational Biology. My research interests lie in the intersection between machine learning on graphs and network science. I’m currently working on temporal graph learning for biology and contributing to the broader efforts in AI 4 Science. I have a Ph.D in Computer Science obtained from the University of Milan, where I worked on machine learning for temporal heterogenous graphs in Connets Lab. Enemies to Lovers story with Python. I’m a former visiting researcher at the University of Edinburgh and Google Developer Student Club lead. Last but not least, I’m a huge Doctor Who fan.

News

• Paper accepted at Neurocomputing! Read “Tensor factorization for temporal knowledge graph forecasting”
• Happy to announce that our work “Enhancing neural link predictors for temporal knowledge graphs with temporal regularisers” has won the best paper award at ESANN 2025!

Latest publications

• Paper accepted at Machine Learning Journal! Read “User migration in blockchain-based online social networks through the lens of temporal node representation shift”
• Paper accepted at TMLR! Read “Evaluating explainability techniques on discrete-time graph neural networks”
• For a full and updated list of my publications see my Google Scholar

Talks

• Tensor Decomposition for TKG reasoning: from completion to forecasting: Talk about tensor decomposition techniques for TKG reasoning held for the Temporal Graph Learning reading group.
• Demystifying Graph Neural Networks: General talk about how GNNs work, when they are successful, and when they can fail, as GNNs do not represent the state-of-the-art for any kind of graphML task.
• Introduction to Graph Neural Networks: Teaching material for machine learning on graphs’ lectures and lab. sessions held for various seminars, master degree and PhD courses on graphML. Recording of a seminar on GNNs is available for those who understand italian.