Research Of Modeling The Seat Preference Of Civil Aviation Passengers And Its Applications

With the rapid development of Internet and e-commence, how to accurately grasp users’preference and demands has increasingly become a hot research problem in the field of electronic commerce. Accurate user profiling and high-efficiency user preference modeling, can not only provide users with personalized service and recommendation, but also improve enterprises’customer relationship management (CRM) and thus increase the benefit of enterprises. Similarly, modeling the seat preference of civil aviation passengers exists in the civil aviation area. Under the trend of economic globalization and aviation federation, civil aviation passenger transport is facing a fiercer and fiercer competition and the enterprises try their best to improve the operational efficiency and thus improve their industrial competitiveness. To civil aviation enterprises, mining out passengers’preference, enhance the understanding of passengers and provide passengers with targeted and personalized service become one of their important means to improve customer relationship management (CRM). For example, airlines try to improve the satisfaction of passengers by allocating seats to them according to their preference.In this paper we study the problem of modeling seat preference of air passengers from the perspective of social networks, taking accessibility of seat resource and social relation between passengers into account, and propose a combined method to model the individual and social preference of passengers.First, we utilize passengers’ historical travel records to infer social relations in the civil aviation area and take these social relations to construct Co-travel networks. Then, based on the fact that seat resource is limited, we present an individual preference model; and based on Co-travel networks, we explore the correlation between the closeness of social relations and the seat distance, and present a social preference model. Finally, we recommend seats to individual passengers based on the proposed preference model and employ the First-Come-First-Serve algorithm to allocate seats for all the passengers in a flight.The experimental results on a real data set in the civil aviation area demonstrate that the proposed preference model can organically combine individual preference with social preference and well describe passengers’preference when travelling with others, which provides a good theoretical support to the seat recommendation problem. What’s more, the experimental results also show that, compared to the seat allocation generated by the check-in system, the seat allocation using the First-Come-First-Serve algorithm improves the passengers’satisfaction.