公交乘客满意度相关因素的影响程度分析
欧阳剑1,胡郁葱2,李橘云1,李健行1

1)广州市交通规划研究院技术咨询所,广东广州 510030; 2)华南理工大学土木与交通学院,广东广州 510640

交通运输系统工程; 常规公交; 乘客满意度; 候车时间; 朴素贝叶斯; 非线性关系

Influence analysis of the factors related to bus passenger gross satisfaction
Ouyang Jian1, Hu Yucong2, Li Juyun1, and Li Jianxing1

1)Technical Consultation Department, Guangzhou Transport Planning Research Institute, Guangzhou 510030, Guangdong Province, P.R.China 2)School of Civil Engineering and Transportation, South China University of Technology, Guangzhou 510640, Guangdong Province, P.R.China

transportation system engineering; regular public transit; passenger gross satisfaction; waiting time; Naive Bayes; non-linear relationship

DOI: 10.3724/SP.J.1249.2016.03288

备注

通过大量乘客满意度问卷调查,利用朴素贝叶斯分类法,建立候车时间、步行时间和乘车环境满意度等9个属性与乘客总体满意度之间的满意度模型,该模型预测精度达79.9%.结果表明,候车时间与乘客总体满意度之间呈非线性关系,当乘客候车时间<5 min的比例<33.8%时,候车时间<5 min的比例每增加1%,乘客总体满意度增加0.07%; 当乘客候车时间<5 min的比例>33.8%时,候车时间<5min的比例每增加1%,乘客总体满意度增加0.04%.

On the basis of a large number of questionnaires, a model is proposed to describe the relationship between passenger gross satisfaction(PGS)and nine properties such as passenger waiting time, walking time, travel conditions and so on by using the Naive Bayes classifier. The model achieves the prediction accuracy of 79.9%. The results show that the relationship between PGS and bus waiting time is nonlinear. If the proportion of bus waiting time, which is less than 5 min(PBWT5), increases by 1 percent, the PGS will increase by 0.07 percent when PBWT5 is less than 33.8%. However, when PBWT5 is over 33.8%, if PBWT5 increases by 1 percent, the PGS will increase by 0.04 percent.

·