基于旅客信任网络的航线选择行为预测

Prediction of route selection behavior based on passenger trust-network

分析理解民航旅客出行特征,对旅客未来潜在出行行为进行预测,是航空公司实施精准营销的重要支撑。该文以民航领域积累的大规模PNR数据集为基础,提出一种基于旅客信任网络的协同过滤航线推荐模型,借鉴社交关系网络引入旅客信任网络,对采用协同过滤进行航线推荐的方法进行改进,通过旅客信任网络中信任的传递性以发现相似旅客,从自身偏好和相似旅客偏好两个方面对旅客出行时对航空公司航线的选择行为进行刻画。实验结果表明,文中算法相较于传统的基于历史选择、基于航线热度等方法有更高的精准度和召回率。

The analyzing and understanding of the travel characteristics of civil aviation passengers and the prediction of their potential travel behaviors in the future are the important supports for precision marketing implementation of airlines.Based on the large⁃scale PNR data sets accumulated in the field of civil aviation,a collaborative filtering route recommendation model based on the passenger trust⁃network is proposed.The social relational network is used to introduce the passenger trust⁃network to improve the method which adopts collaborative filtering to achieve route recommendation.The similar passengers are found by means of the transitivity of trust in the passenger trust network.The passengers′choice behavior for airline routes when they plan to travel are described in two aspects of their own preferences and similar passengers′preferences.The experimental results show that in comparison with the traditional methods based on history selection or route attraction and so on,the algorithm proposed in this paper has higher accuracy and recall rate.