摘要
随着全球定位系统、移动互联网等技术的发展及应用普及,可以获得大量关于乘客的出行轨迹数据,但有关乘客出行目的的信息却难以获得.准确获得乘客出行目的可以更好地服务于乘客(例如基于乘客目的为乘客提供推荐服务).本文采用信息熵理论建立出租车乘客出行目的识别决策树,通过对出租车乘客出行目的的历史数据的分析,将识别结果不正确的元组改正后加入信息库,重新创建决策树,使得该方法具备了一定的学习能力,提高了出行目的识别的准确性.最后本文选取北京地区的出租车乘客数据进行实验,结果表明,该方法能够有效地对出租车乘客出行的目的进行识别.
As the global positioning system（ GPS）,the development and application of mobile technology,such as the Internet popularization,can get a lot of passengers travel track data,but information about passengers travel purpose is hard to come by. Accurate for passenger travel purpose can be better service to passengers（ based on passengers to provide recommendation service for passengers,for example）. Based on the information entropy theory to establish the taxi passenger travel purpose to identify the decision tree,through the taxi passenger travel purpose of the analysis of the historical data,the result will not correct tuples correction after joining information database,to create a decision tree, makes the method have a certain ability to learn, to improve the identification accuracy for travel purpose. Finally this article selects Beijing taxi passenger data experiment,the results show that the method is effective for taxi passenger travel purpose for identification.
作者
潘秀琴
孙东银
PAN Xiu-qin SUN Dong-yin(School of Information Engineer,Minzu University of China Beijing 100081)
出处
《中央民族大学学报（自然科学版）》
2016年第4期41-45,共5页
Journal of Minzu University of China(Natural Sciences Edition)
关键词
信息熵
决策树
出租车乘客
出行目的
识别
Information entropy
Decision tree
Passengers
Travel purpose
Intention recognition
