Skyline-Based Sorting Approach for Rail Transit Stations Visualization

Zhi Cai, Fangzhe Liu, Qiong Qi, Xing Su, Limin Guo, Zhiming DingISPRS Int. J. Geo Inf 2023 · DOI

TL;DR

This paper introduces a novel "Skyline-Based Sorting Approach" to dynamically visualize important rail transit stations on maps. It addresses the issue of cluttered and static station displays by intelligently filtering and ranking stations based on spatial diversity, passenger flow, and transfer status, leading to a more intuitive and user-friendly experience.

What Problem Does This Paper Address?

Urban rail transit systems are becoming increasingly vital for public transportation, and visualizing their stations on maps is crucial for navigation and planning. However, existing map services often suffer from several problems:

  • Information Overload and Occlusion: At various zoom levels and when the map is panned, station icons can overlap, making it difficult to identify individual stations and their information.
  • Static Display: The displayed stations often remain the same even when the map is dragged or zoomed, failing to adapt to the user's current view and needs.
  • Lack of Prioritization: All stations are treated equally, leading to the display of less critical information while potentially obscuring more important ones.

This paper aims to solve these issues by providing a dynamic and intelligent visualization system that prioritizes and displays the most relevant rail transit stations based on user context and station importance.

Methods

The core of the proposed method lies in a "Skyline-Based Sorting Approach" that combines filtering and sorting techniques. The key components are:

  • Three Constraints for Station Importance:
    • Spatial Diversity: Ensures that selected stations are spread out and not clustered in one direction, using a formula based on the relative positions of stations.
  • Time-Sharing Passenger Flow Analysis: Leverages historical passenger data to understand station usage patterns and identify stations with similar or high passenger traffic. This is quantified using an equivalent time-per-hour metric and a similarity calculation.
  • Transfer Station Identification: Assigns a higher importance to stations that serve as transfer hubs, recognizing their critical role in the network.
  • Skyline Query: This multi-objective optimization technique is used to identify "Skyline Points" (SPs) – stations that are not dominated by any other station across the defined constraints.
  • Top-k Sorting: After identifying potential SPs, a weighted scoring function is applied, combining the three constraints with coefficients (α, β, γ). This score determines the "most representative" stations. The algorithm then selects the top-k stations, where 'k' dynamically adjusts based on the map's zoom level.
  • Overlapping Icon Filtering: Before displaying, a check is performed to ensure that the selected top-k stations do not have overlapping icons, further refining the visualization.

Datasets

  • Beijing Subway Station Data: This dataset contains information on 227 subway POIs, including station names and their latitude/longitude coordinates.
  • Beijing Rail Transit System AFC Data: This dataset comprises passenger travel data collected from June to July 2013, including records of passengers entering and exiting stations. This data was used to calculate passenger flow for the 227 stations.

Key Insights and Ideas

  • Dynamic Relevance is Key: The paper highlights that simply displaying all stations is insufficient. True utility comes from dynamically presenting the most relevant stations based on the user's current view and context.
  • Multi-Dimensional Importance: Station importance isn't a single metric. Combining spatial distribution, passenger flow, and transfer status provides a more holistic and accurate assessment.
  • Skyline Meets Top-k: The synergy between Skyline queries (identifying non-dominated points) and Top-k queries (selecting the best 'k' results) is crucial for managing large datasets and ensuring a manageable, yet informative, output.
  • Adaptive Visualization: The concept of 'k' (the number of displayed stations) being dynamically linked to the zoom level is a practical approach to managing visual density and information relevance.

Key Results

  • Improved Visualization: The proposed method effectively reduces icon overlap and provides a clearer display of important stations, especially when the map is dragged or zoomed.
  • Dynamic Updates: The system successfully updates the displayed stations in real-time as the user interacts with the map (panning or zooming).
  • Efficiency: Experimental evaluations showed that the calculation time of the algorithm is positively correlated with the number of POIs and the value of 'k', but remains efficient enough for interactive use.
  • User Satisfaction: User evaluations indicated that the method's displayed key information was consistent with user perception of importance, leading to improved user experience.

Limitations and Future Work

  • Parameter Tuning: The paper acknowledges that the coefficients α, β, and γ play a significant role and that finding optimal values might require further tuning based on specific user preferences or regional characteristics.
  • Future Work: The authors suggest further distinguishing surrounding stations and optimizing the retrieval method to improve overall retrieval efficiency.

Summary updated

BibTeX
@article{cai2023skyline,
  author = {Zhi Cai and Fangzhe Liu and Qiong Qi and Xing Su and Limin Guo and Zhiming Ding},
  title = {{Skyline-Based Sorting Approach for Rail Transit Stations Visualization}},
  journal = {ISPRS Int. J. Geo Inf.},
  year = {2023},
  doi = {10.3390/ijgi12030110},
}
Figure