一种基于社交网络的民航新旅客成长性预测方法

对于一个高效的客户关系管理系统而言，预测客户成长性是必不可少的环节．本文尝试根据新旅客的短期历史出行数据对其未来价值成长性进行预测．为了克服新旅客历史出行记录稀少而导致无法准确预测的缺陷，提出了一种基于旅客社交网络的预测方法．首先根据旅客历史出行记录构建旅客同行网络；然后分另lj从旅客个体和旅客关系的角度构建多种分类特征，用来进行旅客成长性预测；最后提出了一种结合个体预测与关系预测的组合预测模型，以达到准确预测的目的．在某航空公司的真实数据集上进行实验，证明了本文提出的方法对民航新旅客未来价值成长性的推断是有效的．

Predicting the growth of new passengers in civil aviation based on social networks

It is essential to predict the growth of customers for an effective customer relationship man- agement （CRM） system. In this paper, we attempt to predict the future value of new passcngers based on their historical data in a short time. To overcome the rareness of historical data of new pas- sengers, we propose a prediction method based on passenger social networks. First, we construct co- travel networks by extracting social relations between passengers from their historical travel records. Then, we generate a series of individual-based and relation-based features respectively to predict the passenger growth by employing basic classifiers. Finally, we present an effective ensemble model to combine the predictions of individual-based and relation-based classifiers. Experimental results on a real data set of passenger travel records provided by an airline demonstrate that our proposeit approach can efficiently predict the growth of new passengers.