Context-Aware Aviation Auxiliary Services Recommendation Based on Clustering

Abstract

Personalized auxiliary service is an important profit source for airlines. Considering massive passengers’ online and offline travelling context data, the traditional collaborative filtering (CF) algorithm has low accuracy in auxiliary service recommending. This paper adds real-time travel context factors to passenger auxiliary service preference modelling, using five-dimensional data sets to construct a context-aware aviation auxiliary service recommendation model. Aiming to solve the Cold Start and data sparsity, the paper proposes a context-aware recommendation method which calculating similarity between the current context and historical context other than passenger’s reviews on services. And recommends the target passenger top-N auxiliary service items that under historical similar travel context. Finally, the experimental simulation method is used to evaluate the recommended effect and accuracy of the recommendation system. The result shows that the context-aware recommendation method is higher accurate in personalized aviation auxiliary services recommending.