Eric Piette
As an Assistant Professor affiliated with UCLouvain in the Information and Communication Technologies, Electronics and Applied Mathematics (ICTEAM) institute and the Computing Science and Engineering (INGI) department, my primary focus is on Game Artificial Intelligence (AI) and its applications.
I specialize in various areas such as Machine Learning (ML), Explainable AI (XAI), Constraint Satisfaction Problem (CSP) and Knowledge Representation and Reasoning (KRR) in the context of General Game Playing (GGP).
I am the main proposer and Chair of the GameTable COST Action CA22145 - Computational Techniques for Tabletop Games Heritage. An international and interdisciplinary network of scholars and stakeholders from all career stages across academia, industry, and heritage institutions to inspire methodologies and applications on how to use game AI to study, reconstruct, and preserve the intangible cultural heritage of games.
I was previously part of the Digital Ludeme Project, a five-year ERC-funded project running from April 2018 to August 2023. The project aimed to provide a computational study of traditional strategy games throughout recorded human history and explored their role in human culture and the spread of mathematical ideas. In this project, we developed the Ludii General Game System, which is designed to play, evaluate and design a wide range of games such as board games, dice games, mathematical games, and more. Using structured sets of ludemes (units of game-related information), Ludii can model the full range of traditional strategy games from around the world in a single playable database.
My Ph.D. thesis entitled "A Stochastic Constraint-Based Approach to General Game Playing" introduced a new approach based on the stochastic constraint satisfaction problem (SCSP) and the Bandit-based stochastic sampling for General Game Playing. This approach simplifies the resolution of each problem by demonstrating the equivalence between the detected constraint/solution symmetries in the constraint network generated with the structural/strategies symmetries described by the GGP games. In practice, this work is substantiated by the general game player called WoodStock, which won the International General Game Playing Competition 2016 (IGGPC'16) organised by Stanford University.
Keywords: General Game Playing, Ludii, Machine Learning, Knowledge Representation, Constraint Programming.
Artificial Intelligence is no substitute for natural stupidity.