RTSKG: Building a Rail Transit Station Knowledge Gr...

RTSKG: Building a Rail Transit Station Knowledge Gr... | AI Research

Key Takeaways

RTSKG: Building a Rail Transit Station Knowledge Graph Dataset introduces a structured dataset designed to help researchers analyze urban rail transit system...
Rail transit systems play a vital role in urban mobility and economic development.
As key components of such systems, rail transit stations function as critical transport hubs that enhance urban accessibility and stimulate development in surrounding areas.
RTSKG integrates heterogeneous urban entities, such as rail transit stations, road segments, and points of interest, with a specially designed unified schema, and is accessible as Linked Data at this https URL .
Evaluations on station-area store recommendation and knowledge-enhanced ridership prediction demonstrate the effectiveness of RTSKG, highlighting its potential to support city-level rail transit station analysis.

Paper AbstractExpand

Rail transit systems play a vital role in urban mobility and economic development. As key components of such systems, rail transit stations function as critical transport hubs that enhance urban accessibility and stimulate development in surrounding areas. City-level rail transit station related tasks (e.g., ridership prediction) require large-scale urban data, but current studies often neglect complex interactions among various urban entities in terms of data organization. In this paper, to address the above issue, we build a Rail Transit Station Knowledge Graph (RTSKG) dataset which explicitly models the spatial and semantic interactions among different kinds of urban entities, to benefit city-level rail transit station related tasks. RTSKG integrates heterogeneous urban entities, such as rail transit stations, road segments, and points of interest, with a specially designed unified schema, and is accessible as Linked Data at this https URL . Evaluations on station-area store recommendation and knowledge-enhanced ridership prediction demonstrate the effectiveness of RTSKG, highlighting its potential to support city-level rail transit station analysis.

RTSKG: Building a Rail Transit Station Knowledge Graph Dataset introduces a structured dataset designed to help researchers analyze urban rail transit systems. By explicitly modeling the spatial and semantic relationships between rail stations and their surrounding urban environment, the dataset aims to improve the accuracy of city-level tasks like ridership prediction and store recommendation.

Addressing Data Fragmentation

Current urban analysis often treats rail stations, roads, and points of interest (POIs) as independent entities. This approach fails to capture how these elements influence one another—for example, how the proximity of a POI to two different stations might cause those stations' ridership numbers to fluctuate in relation to each other. The authors, Shutong Zhu et al., developed RTSKG to bridge this gap by integrating heterogeneous urban data into a unified knowledge graph that maps these complex interactions.

Dataset Construction

The researchers constructed RTSKG using data from New York City and Chicago, covering four primary categories: rail transit, administrative divisions, road networks, and POIs. The team defined a custom ontology consisting of ten classes—such as "Station," "Station Area," "Block," and "Functional Area"—and nineteen distinct relations. These relations categorize how entities interact, including geographic location, adjacency, and intersection. To define "Station Area," the team used the Mapbox Isochrone API to calculate the regions reachable from station entrances within 5-minute and 10-minute walking thresholds.

Knowledge Graph Embedding

To make the data useful for predictive tasks, the authors applied various Knowledge Graph Embedding (KGE) models to convert the graph’s instances and relations into vector representations. After testing eleven different models, the researchers found that GIE (Geometric Interaction Embedding) performed best. They attribute this success to GIE’s ability to simultaneously capture the hierarchical structures (such as the relationship between stations, functional areas, and boroughs) and the cyclic structures (such as the overlap between station areas, POIs, and blocks) present in the RTSKG ontology.

Practical Utility

The authors evaluated the dataset through two specific tasks: station-area store recommendation and knowledge-enhanced ridership prediction. By leveraging the structured relationships within RTSKG, these tasks achieved better performance compared to using existing urban knowledge graphs. The dataset is currently available as Linked Data, providing a standardized resource for future research into transit-oriented development and urban mobility.