The Analysis And Research On Civil Aviation Passengers' Behavior And Value

In today's era, the civil aviation industry rapidly develops, which causes high- quality customer resources become an important way to measure the competitiveness between airlines, and competitions for high-value customers become the core way of airline profitable.With the help of research on customer value measurement model, airlines could analysis civil aviation passenger value from quantitative point of view. And airlines can identify high-value,medium-value and low-value civil aviation passengers by using cluster analysis method. All of this helps airlines to use the least investment of resources to obtain the maximum profit return by allocating different resources to customer at different customer value levels.The information of civil aviation passengers' value is contained in a series of passengers' behavior record data which are produced in the process of airline passengers' booking,check-in, flight, after flight and so on. By using customer behavior records data, which contains the following attributes: reservations way, flight routes, mileage, accommodation,fares, payment, passengers' habits, special requirements, feedback, complaints, etc, this paper presents a model of civil aviation passengers' value measurement based on data mining methods. And by using this model, we can recognize different levels of customers such as excellent ones, ordinary ones and potential ones, and can provide theoretical support for the airlines' differential marketing strategies.Works in this paper are mostly focused on the following topics:Considering that civil aviation passenger behavior information is often contained in the unstructured data, this paper proposed a quick 2-5 word indexing segmentation algorithm to extract information from user behavior data. And then this segmentation algorithm is integrated into the Lucene system so as to get user behavior information fast. Experiment shows that this segmentation algorithm has better time efficiency and segmentation accuracy.By using user behavior information, the factors of civil aviation passengers' value are constructed and civil aviation passengers' value model is proposed. In the process of constructing civil aviation passengers' value factors, this paper proposes a method based on principal component analysis. With the help of this method, the mutual relationship between civil aviation passengers' value indicators is removed, and reduction on the factors' dimensions is more conducive to calculations. And in the study of civil aviation passengers' value, this paper proposed two models based on principal component analysis and fuzzy analytic hierarchy process.Based on k-means clustering method, this paper proposed a civil aviation passengers' value clustering method by using the civil aviation passengers' value model constructed above.By using this civil aviation passengers' value clustering method, different score range such as high, medium and low customer value range are obtained. Finally, based on civil aviation passengers' value, the paper gives a differential marketing strategy for airlines, and this strategy provides theoretical support for airlines to make marketing strategies and ehance service quality.