隨著科技的快速發展與進步,為提高出門的便利性,各式各樣交通工具的出現提供給社會大眾不同的選擇,由於計程車的便利性與可達性,在整個交通運輸上扮演著關鍵的角色。因此,為提高計程車駕駛工作效率有效的資訊,載客行為的研究是一個重要的課題。 本文使用滯後序列分析找出台北市地區之計程車駕駛人之載客路線轉移傾向,並透過 K-pprototypes以及廣義相關圖(GAP)將計程車駕駛人根據其屬性分成「集中載客型」與「混合載客型」兩群,「集中載客型」具年資低、工作時數較長、車齡較短等特徵;「混合載客型」具年資高、工作時數較短、車齡較長等特徵。 不同計程車屬性資料與轉移類型間存在顯著差異。「派遣車隊」在跨區轉移中的比例顯著的高於「非派遣車隊」,「非派遣車隊」在區內轉移中的比例顯著的高於「派遣車隊」;「車行計程車」在跨區轉移中的比例顯著的高於「合作社計程車」及「個人計程車」,「合作社計程車」和「個人計程車」在區內轉移中的比例顯著的高於「車行計程車」。
With rapid advances of technology, various transportation modes have sprung up to improve our travel convenience, and amongst taxis play as a crucial role in transportation service due to their convenience and accessibility. How to mine driving behaviors of taxi drivers is thus an important issue that offers beneficial information to improve drivers’ work efficiency. In this study, we first employ the so-called lag sequential analysis to explore taxi drivers’ operation pattern. Subsequently, K-prototypes clustering algorithms along with generalized association plots are utilized to provide a deeper understanding about the driving behaviors based on drivers’ personal attributes and their cab features. The analysis results show that taxi drivers in Taipei city can be classified into two types. One is the "concentrated drivers" who are of shorter driving tenure, long working hours, and driving an older vehicle. The other is the "mixed drivers" who own a younger cab and have personal characteristics of longer seniority and shorter working hours. In addition, there exist significant differences between taxi drivers' attribute and operation pattern. Drivers from company with taxi dispatching service travel between districts most often than those without taxi dispatching service. The same observation can be found between drivers who have been members of a taxi company and those who cooperative as a taxi driver. Finally, the possibility of traveling within districts is significantly lower for driver members of a taxi company than those who are cooperative as a taxi driver or with private hire vehicles.