A Study on License Management of Taxi Drivers Using Data Mining Analysis ~ A Case Study on Taipei Metropolitan Area

• Main Authors: Han Chen Wu, 吳漢城
• Other Authors: C. H. Chen
• Format: Others
• Published: 2008
• Online Access: http://ndltd.ncl.edu.tw/handle/99565935618742291455

碩士 === 長庚大學 === 資訊管理學研究所 === 96 === There are not many researches on Taxi industry in Taiwan, even though there are some reports emphasizing on taxi operation and security. This research focuses on the analysis of the population structure and the prediction for their future license status. This research analyses the database of license management for taxi drivers in Taipei metropolitan area with data mining and statistical methods. From the analysis, the fllowings are observed. (1)On the average, the age of taxi drivers is getting younger, and their educational level is getting higher. Therefore, the education level of a young taxi driver is higher than that of an old taxi driver on the average. (2)By only using driver data from discriminative attributes to create classification models, we obtain better accuracy. (3)When the age, education level and birthplace data of taxi drivers are used to predict tutorial examination result and the reason for license cancellation of taxi drivers, the accuracies are 77.21% and 63.56% respectively. (4)When the age, birth season and education data of taxi drivers are used to predict service seniority of taxi drivers, the accuracy is 53.99%. (5)When the age, birth season, sex, education and birthplace data of taxi drivers are used to predict the reason for violation of license registry, the accuracy is 84.74%.