Exploring the Service Quality of Public Bike-Sharing System Using Data Mining Methods

碩士 === 國立臺北科技大學 === 工業工程與管理系碩士班 === 104 === Cities around the world have put efforts on pollution, carbon reduction, traffic congestion and public health issues. These wide ranges of the growing public services lead to municipal tight budgets. Therefore, public bike sharing system gradually becomes...

Full description

Bibliographic Details
Main Authors: Chang Yi-Wu, 吳昌益
Other Authors: 劉建浩
Format: Others
Language:zh-TW
Published: 2016
Online Access:http://ndltd.ncl.edu.tw/handle/y4nq5v
Description
Summary:碩士 === 國立臺北科技大學 === 工業工程與管理系碩士班 === 104 === Cities around the world have put efforts on pollution, carbon reduction, traffic congestion and public health issues. These wide ranges of the growing public services lead to municipal tight budgets. Therefore, public bike sharing system gradually becomes the most popular municipal solution. From the green city point of view, this bike-sharing system may give cities a new life. Prior studies have shown that good public transport service quality can attract more passengers to use public transportation. With the popularity of public bike-sharing systems around the world, many relevant studies have shown nowadays. However, most previous studies focused on the implementation of the system, site selection and impacts on the other vehicle. This study aims to discuss how to improve the service quality public bike-sharing system. First, this study built up a system with four dimensions and 14 criteria on Taipei You Bike system by the literature review and interviews with experts. Using the data mining techniques “Random Forest (RF)”, we derived some key factors relative to service quality. Then, the Dominance-based Rough Set Approach (DRSA) was used to analyze the surveyed data. A set of "if ..., then ..." decision rules can help decision-makers understand the key factors that users perceived. We further used the flow graphs to visualize these rules and helps decision-makers understand more easily. Finally, based on the analyzed results, we provided some suggestions for improving the service levels of public bike-sharing system.