The PILGRIM (ProbabIListic Graphical and RelatIonal Models) library aims at providing efficient software tools in order to define and learn different probabilistic graphical models (currently directed ones: Bayesian networks, Probabilistic Relational Models, Graphical Event Models), and use them to make inference on data. It is composed of 3 different projects:
- PILGRIM General allows to define, learn and serialize Bayesian networks (including Dynamic Bayesian networks [Trabelsi 2013]). It offers structure learning algorithms for Bayesian networks (e.g. greedy search, MMHC),incremental structure learning [Yasin 2013], structure evaluation (e.g. BDeu, BIC, qNML, SHD) and multi-task learning [Benikhlef 2025]
- PILGRIM Relational offers many features for the definition and learning of Probabilistic Relational Models (PRM) and their extensions, as well as their instantiation into grounded models for inference [Coutant 2015, Ben Ishak 2015, Elabri 2018].
- PILGRIM Evential allows to define, learn and serialize Graphical Event Models. It includes structure learning algorithm (greedy search) [Said Hawchar 2025] as well as approximate probabilistic inference algorithms [Hribach et al. 2026].
PILGRIM is developed and maintained by members of the Data, User, Knowledge (DUKe) team of the CNRS LS2N Laboratory, in Nantes, France, with the help of Ouest Valorisation, Technology transfer acceleration company. The project is led by Pr. Philippe Leray.
ProBT dependency
PILGRIM General and Relational projects encapsulate the ProBT library, developed by ProbaYes. ProBT allows to model, learn and make inference in Bayesian networks and PILGRIM reuses many of its classes defining basic concepts in probabilistic programming, while extending its capabilities in terms of Bayesian networks structure learning and in terms of relational datasets support.
Application Layers
- LTEC (Lifetime Expense Curve), a model of the aging of a system’s components and the evolution of its maintenance costs [Blanchard et al. 2023]
- BKT (Bayesian Knowledge Tracing), modeling and monitoring a learner’s skills over time [Couland et al. 2023]
- RecSys (Recommender System): Building recommendation models based on probabilistic relational models [Chulyadyo 2016]
Example of applications in specific projects
- Outlier detection in tabular data (project with Talend) [Dufraisse et al. 2020]
- Modeling the aging of wind farm components, diagnostic support (FUI Hyperwind project) [Gherasim et al. 2016]
- Simulation of real estate affordability (EP collaborative project) [Blanchard et al. 2023]
- Support for preventing repeat suicide attempts (Medical Companion project with Nantes University Hospital) [Mouchabac et al 2021]
- Tool for grading viper venom toxicity (Vipgrade project, with Nantes University Hospital)[Mazzobel et al. 2021]
- Modeling of surgical procedures (Cominlabs SPARS project) [Said Hawchar 2025]
- Scoring and recommendations for operational optimizations in ship navigation (project with Ecomer Data)