A Graph Machine Learning Approach for Classifying Museums Layouts Based on their Circulation Patterns and Gallery Experiences.

This project aims to use graph machine learning to evaluate and classify museum layouts and experiences. By leveraging the structure and relationships within a museum’s most contributing elements, such as galleries and circulation. The project aims to identify patterns that give rise to different museum experiences. Graph-machine learning and namely graphSAGE frameworks for representation learning are employed.

The project initially studies and classifies the gallery spaces – or nodes – based on their connectivity. Then, through a more zoomed out lens, museoflow tries to establish a system of classifying museums as a whole.