Towards eXplainable AI for Mobility Data Science
This work targets the problem of making AI models interpretable for mobility data applications, such as GPS tracking, but it is incremental as it builds on existing GeoXAI studies and outlines a research path without new findings.
The paper addresses the need for explainable AI in mobility data science by proposing the use of temporal graph neural networks and counterfactuals to learn from dense trajectory data, but it does not present specific results or concrete numbers as it is ongoing work.
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.