ANALYZING DIVERSITY IN BIKE-SHARING USAGE PATTERNS

Abstract

Bike-sharing systems have been introduced in various cities worldwide, and understanding their usage patterns is crucial for improving the service. This study analyzes bike-sharing usage data from Kumamoto City, Japan, to 1) identify the characteristics of rides where the departure and arrival ports are the same, and 2) reveal the diversity of rides using an index based on the Gini coefficient. We propose four diversity indices: departure time diversity, port pair diversity, departure port diversity, and arrival port diversity. These indices, along with the Lorenz curve, are used to visualize and quantify usage diversity. The results indicate a decreasing trend in departure and arrival port diversity as the distance from downtown Kumamoto increases.