Resources.
The inspiration for our project is a paper presented at the 2019 Intelligent User Interfaces conference (see link below). In this paper, the authors attempt to create an AI that can autonomously provide human-readable explanations to complement decisions made by another AI playing the game of the frogger. We implemented our own decision generator for Pac-Man from a combination of ideas we read about in this paper and our own original strategies.
Automated Rationale Generation: A Technique for Explainable AI and its Effects on Human PerceptionsWe adapted a pre-built Pac-Man game written in Python 2.7, provided by the Berkeley CS Department.
Berkeley Pac-Man Codebase