Discovering Abnormal Civil Aviation Requirements By Analyzing Online Flight Ticket Query Behaviors

In today's civil aviation market,accurate grasp of passenger requirements is the core issue of airline revenue management.The continuous growth of the civil aviation market put forward higher requirements in terms of business management and operational level,especially the market response capabilities to airlines,airports,air tickets agents and other civil aviation related.This requires the civil aviation enterprise can promptly and accurately grasp the change of civil aviation market demand,market and take corresponding countermeasures in a timely manner,so as to improve the enterprise's operational capacity and service quality,increase revenue,and improve the user experience of travel.The changes of users' query volume in online fight ticketing systems indicate the changes of requirements in civil aviation market.In this paper,we analyze the big data of users' online query behaviors,which can timely and accurately discovery abnormal civil aviation requirements.It is very conducive for airlines and agencies to take effective marketing actions immediately.Firstly,for the problem of the abnormal value on each sliding window in single route query time series.This paper presents a method for calculating the abnormal value from multiple dimensions,respectively:the difference between this window of the route query and the window with the history query volume curve over the same route and same period,the difference between this window of the route query and the window in other route curve at the same time and the window of the route query the amount of time complexity of sequence curve.Finally,the initial value of the outliers of sliding window sub sequences are obtained by combining the outliers calculated from these three dimensions.Secondly,we built an airline route network from the city owned airport and airline relations.Based on the network,this paper proposes a demand outliers network iterative optimization algorithm.We iteratively adjust and optimize the demand anomaly from the perspective of the whole network,until finally achieve a balance on the entire route network,get the final route anomaly.We carry out experiments on a real-world users' query dataset collected from an GDS service providers and compared with traditional time series anomaly found algorithm.The experimental results demonstrate that the proposed method can effectively discovery abnormal civil aviation requirements from users' online query logs.Finally,we apply the method of anomaly demand discovery in the civil aviation demand index system,and use the results of anomaly discovery to study the related work.