Summary of Heterogeneous Graph Neural Networks with Post-hoc Explanations For Multi-modal and Explainable Land Use Inference, by Xuehao Zhai et al.
Heterogeneous Graph Neural Networks with Post-hoc Explanations for Multi-modal and Explainable Land Use Inference
by Xuehao Zhai, Junqi Jiang, Adam Dejl, Antonio Rago, Fangce Guo, Francesca Toni, Aruna Sivakumar
First submitted to arxiv on: 19 Jun 2024
Categories
- Main: Artificial Intelligence (cs.AI)
- Secondary: None
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The paper introduces an explainable framework for inferring land use from multi-modal mobility data, which synergizes heterogeneous graph neural networks (HGNs) with Explainable AI techniques to enhance both accuracy and explainability. The proposed HGNs significantly outperform baseline graph neural networks for six land-use indicators, particularly for ‘office’ and ‘sustenance’. Feature attribution explanations show that the framework predicts symmetrical patterns of ‘residence’ and ‘work’ categories aligned with commuter activities in London, while counterfactual explanations reveal that variations in node features and types drive differences between predicted land use distributions and ideal mixed states. The proposed HGNs can support urban stakeholders in their planning and policy-making decisions. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The paper uses sensor and location technologies to collect multi-modal mobility data for daily activity patterns. It then introduces an explainable framework that combines heterogeneous graph neural networks (HGNs) with Explainable AI techniques to predict land use. The results show that the proposed HGNs are more accurate than previous methods, especially for ‘office’ and ‘sustenance’ categories. The explanations provided demonstrate how the framework can be used to support urban planning and policy-making. |
Keywords
* Artificial intelligence * Multi modal
