项目名称: 基于集成异构网络的民航旅客-航班关联挖掘研究
项目编号: No.61502499
项目类型: 青年科学基金项目
立项/批准年度: 2016
项目学科: 计算机科学学科
项目作者: 卢敏
作者单位: 中国民航大学
项目金额: 18万元
中文摘要: 民航旅客-航班关联预测和分析能为航空公司新航班的开辟和客户关系管理等提供重要依据。传统方法只通过旅客社交网络进行挖掘,忽略了旅客-航班关联等结构信息。本课题将旅客、航班和旅客-航班关联三个网络集成为一个异构网络,以保留各网络的拓扑信息,在基础之上开展异构网络子图抽取、关联预测和聚类分析等工作:1)将异构网络子图抽取建模为图匹配问题,以抽取面向不用应用需求的子图;2)将旅客聚类和航班聚类的一致性约束引入到旅客和航班的协同聚类中,为潜在高价值新航班挖掘等任务提供分析工具;3)将旅客-航班关联预测建模为优化问题,通过设计目标函数和优化算法的以适应旅客-航班关联稀少问题。
中文关键词: 民航旅客订票日志;集成异构网络;民航旅客-航班关联;数据挖掘;大数据
英文摘要: Predicting and analyzing passenger-flight association could help airlines make better decisions in coming up new flights, managing customer relationship and so on. The traditional algorithms were built on passengers’ social network with complete ignorance of topologies in flight phenotype network and passenger-flight bipartite network. To preserve the topologies of original networks, our proposal integrates the three above networks into one heterogeneous network. Based on it, 1) subgraphs are extracted from the integrated heterogeneous network in the need of applications; 2) a clustering analysis tool is designed to cluster passengers and flights respectively with the consistent constraint from prior information; 3) the problem of predicting passenger-flight association is formulated as an optimization problem, which is solved by bi-random walk. The bi-random walk avoids the bias from sparse associations.
英文关键词: passenger name records;integrated heterogeneous networks;passenger-flight association;data mining;big data