Spatiotemporal Traffic Accident Records from Chinese Court Judgments

Lei Wang's dataset contains approximately 1.57 million records of traffic accidents in China, extracted from national court judgments using large language models. Each record provides precise geographical coordinates (longitude and latitude), a timestamp, and fatality outcomes. It offers a high-resolution spatiotemporal infrastructure for interdisciplinary research on road safety and urban governance.

Use Cases

• Analyzing geographic hotspots of traffic accidents based on longitude and latitude coordinates.
• Studying temporal patterns of accident severity using timestamps and fatality outcomes.
• Modeling the relationship between road infrastructure and accident rates for urban planning.
• Supporting public health research on injury prevention and progress toward SDG Target 3.6.
• Validating or supplementing official traffic statistics with court-derived incident data.

Strengths

• Contains a large-scale sample of approximately 1.57 million accident records.
• Provides high-precision spatiotemporal data with geographic coordinates and timestamps.
• Derived from an authoritative public source: national court judgment records in China.
• Uses advanced LLM techniques for extraction from complex, unstructured judicial texts.

Limitations

• The exact column schema and full variable list are not provided in the available metadata.
• The dataset's temporal coverage (start and end years) is not explicitly stated.
• Relies on court records, which may not capture all accidents (e.g., unreported or settled out of court).

Provenance

Source

Publicly accessible national court judgment records in China.

Collection Method

Extraction from unstructured judicial texts using large language model techniques.

Time Range

null

Freshness

Last updated on 2026-03 24.

Geography

China