mlsanityData to 14 Oct 2026

Learning A Universal Crime Predictor with Knowledge-guided Hypernetworks

arXiv 2511.02336 ↗First posted 4 Nov 2025cs.LGcited by 0

Not on the map. Why?

Abstract (arXiv)

Predicting crimes in urban environments is crucial for public safety, yet existing prediction methods often struggle to align the knowledge across diverse cities that vary dramatically in data availability of specific crime types. We propose HYpernetwork-enhanced Spatial Temporal Learning (HYSTL), a framework that can effectively train a unified, stronger crime predictor without assuming identical crime types in different cities' records. In HYSTL, instead of parameterising a dedicated predictor per crime type, a hypernetwork is designed to dynamically generate parameters for the prediction function conditioned on the crime type of interest. To bridge the semantic gap between different crime types, a structured crime knowledge graph is built, where the learned representations of crimes are used as the input to the hypernetwork to facilitate parameter generation. As such, when making predictions for each crime type, the predictor is additionally guided by its intricate association with other relevant crime types. Extensive experiments are performed on two cities with non-overlapping crime types, and the results demonstrate HYSTL outperforms state-of-the-art baselines.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 29 of its 53 references that are in the map. First posted on arXiv 4 Nov 2025.

Builds on

29 of 53 references.

References · 29

  1. 2017Spatio-Temporal Graph Convolutional Networks: A Deep Learning Framework for Traffic Forecastingcited by 1,136
  2. 2019Graph WaveNet for Deep Spatial-Temporal Graph Modelingcited by 829
  3. 2020Connecting the Dots: Multivariate Time Series Forecasting with Graph Neural Networkscited by 566
  4. 2019EvolveGCN: Evolving Graph Convolutional Networks for Dynamic Graphscited by 447
  5. 2019GMAN: A Graph Multi-Attention Network for Traffic Predictioncited by 434
  6. 2018T-GCN: A Temporal Graph ConvolutionalNetwork for Traffic Predictioncited by 401
  7. 2020Wiki-CS: A Wikipedia-Based Benchmark for Graph Neural Networkscited by 260
  8. 2024UniST: A Prompt-Empowered Universal Model for Urban Spatio-Temporal Predictioncited by 87
  9. 2024UrbanGPT: Spatio-Temporal Large Language Modelscited by 81
  10. 2020A3T-GCN: Attention Temporal Graph Convolutional Network for Traffic Forecastingcited by 54
  11. 2022Spatial-Temporal Hypergraph Self-Supervised Learning for Crime Predictioncited by 39
  12. 2022Spatial-Temporal Sequential Hypergraph Network for Crime Prediction with Dynamic Multiplex Relation Learningcited by 32
17 more
  1. 2022Multi-Graph Fusion Networks for Urban Region Embeddingcited by 32
  2. 2020Generative Adversarial Zero-Shot Relational Learning for Knowledge Graphscited by 30
  3. 2021HAGEN: Homophily-Aware Graph Convolutional Recurrent Network for Crime Forecastingcited by 22
  4. 2023UUKG: Unified Urban Knowledge Graph Dataset for Urban Spatiotemporal Predictioncited by 20
  5. 2023Explainable Spatio-Temporal Graph Neural Networkscited by 13
  6. 2023Spatio-Temporal Meta Contrastive Learningcited by 11
  7. 2023Crime Prediction Using Machine Learning and Deep Learning: A Systematic Review and Future Directionscited by 8
  8. 2024Multi-scale Traffic Pattern Bank for Cross-city Few-shot Traffic Forecastingcited by 7
  9. 2024Hate Speech Detection with Generalizable Target-aware Fairnesscited by 7
  10. 2024Start from Zero: Triple Set Prediction for Automatic Knowledge Graph Completioncited by 7
  11. 2020AIST: An Interpretable Attention-based Deep Learning Model for Crime Predictioncited by 5
  12. 2023Semantic-Fused Multi-Granularity Cross-City Traffic Predictioncited by 4
  13. 2024Frequency Enhanced Pre-training for Cross-city Few-shot Traffic Forecastingcited by 4
  14. 2024Physics-guided Active Sample Reweighting for Urban Flow Predictioncited by 3
  15. 2022Spatial-Temporal Meta-path Guided Explainable Crime Predictioncited by 2
  16. 2024Target inductive methods for zero-shot regressioncited by 1
  17. 2024Memory-enhanced Invariant Prompt Learning for Urban Flow Prediction under Distribution Shiftscited by 1

Built on by

Not cited yet.