Towards eXplainable AI for Mobility Data Science

TL;DR

This paper discusses developing explainable AI models for mobility data science, emphasizing human-centered explanations for complex trajectory data using temporal graph neural networks and counterfactuals.

Contribution

It introduces a research framework for applying XAI techniques to mobility data, integrating temporal GNNs and counterfactual explanations for better interpretability.

Findings

Review of existing GeoXAI studies

Proposed human-centered explanation approaches

Outline of future research directions

Abstract

This paper presents our ongoing work towards XAI for Mobility Data Science applications, focusing on explainable models that can learn from dense trajectory data, such as GPS tracks of vehicles and vessels using temporal graph neural networks (GNNs) and counterfactuals. We review the existing GeoXAI studies, argue the need for comprehensible explanations with human-centered approaches, and outline a research path toward XAI for Mobility Data Science.