民航旅客团体出行目的推断研究与应用

发布时间：2018-06-28 11:09

本文选题：民航旅客团体 + 出行目的　；参考：《北京交通大学》2014年硕士论文

【摘要】：在现实生活中,人们常常因为拥有相同的出行目的地而组成一个旅客团体共同乘机出行。推断民航旅客团体出行目的是一个非常有趣且有价值的研究问题,因为推断结果能为各个航空公司、航空管理部门甚至政府的服务及决策提供支撑,从而带来有意义变化。在本文中,我们尝试在乘客的旅行历史记录数据基础上通过设计分类算法来解决这个问题。为克服传统分类器的独立同分布(IID)假设的约束,本文提出一种基于迭代分类的新型集体分类方法。首先,我们从民航相关部门提供的加密旅客出行记录中抽取出旅客间的社会关系,并基于这些关系构建民航旅客共同出行网络。在共同旅行网络基础上,进一步构造了一组特征指标用于描述不同类别的民航旅客团体,然后使用这些特征指标对旅客团体进行初步预分类。在预分类以后,我们使用集体分类的思想,利用不同旅客团体之间的重叠部分形成类别标签,再使用这个标签不断进行迭代分类,当结果稳定时迭代终止。我们在民航乘客的旅行记录真实数据集上进行了实验,结果表明集体迭代分类方法可以有效地推断民航旅客团体旅行目的。最后,我们将旅客团体出行目的推断方法应用于不同城市出发达到旅客团体的出行目的推断,并利用推断结果有效地分析城市的性质与发展趋势。
[Abstract]:In real life, people often form a group of passengers to travel together because they have the same destination. It is a very interesting and valuable research question to infer the purpose of civil aviation passenger group travel, because the inferred result can provide support for the service and decision of various airlines, aviation management departments and even the government, thus bring about meaningful change. In this paper, we try to solve this problem by designing a classification algorithm based on passenger travel history data. In order to overcome the constraints of the IID hypothesis of traditional classifiers, a new collective classification method based on iterative classification is proposed in this paper. First of all, we extract the social relations between passengers from the encrypted passenger travel records provided by civil aviation departments, and construct the common travel network of civil aviation passengers based on these relationships. On the basis of the common travel network, a set of characteristic indicators is constructed to describe different types of civil aviation passenger groups, and then these characteristics are used to pre-classify the passenger groups. After pre-classification, we use the idea of collective classification to form category labels by using overlapping parts between different passenger groups, and then use this label to iterate and classify, and then terminate iteratively when the results are stable. We have carried out experiments on the real data set of civil aviation passengers' travel records. The results show that the collective iterative classification method can effectively infer the purpose of air passenger group travel. Finally, we apply the method of passenger group travel purpose inference to the travel purpose inference of different cities, and use the inference results to analyze the nature and development trend of the city effectively.
【学位授予单位】：北京交通大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：F592;TP311.13

【共引文献】