Research map / Probabilistic and scientific modeling / Molecules and proteins
SpatialRank: Urban Event Ranking with NDCG Optimization on Spatiotemporal Data
BANG AN, Xun Zhou, Yongjian Zhong, Tianbao Yang
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use learned beta instead of fixed beta=0.5 in graph convolution (ablation) | SpatialRank · Chicago Accident | metric 1 @K=30: .250 -> .257 | Improved |
| use learned beta instead of fixed beta=0.5 (ablation) | SpatialRank · Chicago Crime | metric 1 @K=30: .361 -> .373 | Improved |
| use learned beta instead of fixed beta=0.5 (ablation) | SpatialRank · Iowa Accident | metric 1 @K=30: .530 -> .540 | Improved |
| replace cross-entropy loss with the hybrid NDCG loss and importance sampling | SpatialRank · Chicago Accident | metric 1 @K=30: .232 -> .257 | Improved |
| replace ApproxNDCG with SpatialRank optimization | SpatialRank · Chicago Accident | metric 1 @K=30: .240 -> .257 | Improved |
| replace SONG with SpatialRank optimization | SpatialRank · Chicago Accident | metric 1 @K=30: .240 -> .257 | Improved |
| NDCG-based losses instead of cross-entropy | network · Chicago Crime | metric 1 @K=30: .362 -> .373 | Mixed |
| add local NDCG term to loss (sigma 0.0 -> 0.1) (ablation) | hybrid loss · Chicago Accident | metric 1 @K=30: 0.251 -> 0.256 | Improved |
| add local NDCG term to loss (sigma 0.0 -> 0.1) (ablation) | hybrid loss · Chicago Crime | metric 1 @K=30: 0.366 -> 0.371 | Improved |
| increase sigma from 0.1 to 0.3 (ablation) | hybrid loss · Chicago Accident | metric 1 @K=30: 0.256 -> 0.234 | Hurt |
| increase sigma from 0.1 to 0.3 (ablation) | hybrid loss · Chicago Crime | metric 1 @K=30: 0.371 -> 0.345 | Hurt |
| add importance-based location sampling weights (ablation) | SpatialRank · Chicago Accident | metric 1 @K=30: 0.255 -> 0.257 | Improved |
Closest papers by meaning
- Spatial-Temporal Augmentation for Crime Prediction (Student Abstract) (AAAI 2024 · similarity 0.86)
- GeoPro-Net: Learning Interpretable Spatiotemporal Prediction Models Through Statistically-Guided Geo-Prototyping (AAAI 2025 · similarity 0.86)
- Towards Dynamic Spatial-Temporal Graph Learning: A Decoupled Perspective (AAAI 2024 · similarity 0.83)
- GeoRanker: Distance-Aware Ranking for Worldwide Image Geolocalization (NeurIPS 2025 · similarity 0.83)
- TraffiDent: A Dataset for Understanding the Interplay Between Traffic Dynamics and Incidents (NeurIPS 2025 · similarity 0.82)
- Decision-aware Training of Spatiotemporal Forecasting Models to Select a Top-K Subset of Sites for Intervention (ICML 2025 · similarity 0.82)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
