Cloud based airport customer behavior prediction and personalized service design based on deep learning
In order to improve the efficiency of aviation operation management,research is conducted from the perspective of cloud based airport customer behavior analysis.An improved support vector regression model is selected to predict the airport passenger travel index,and a first come,first served algorithm is chosen for seat recommendation and allocation.The results show that when the number of flights is 525,the overall average satisfaction of the first come first served algorithm is greater than 0.800,which is higher than the historical seat allocation results.The management efficiency score in cloud mode is 0.75,which is 0.22 higher than that in non cloud mode.The satisfaction level in cloud mode is 0.24 higher than that in non cloud mode.The research method can effectively predict the passenger travel index and promote the overall average satisfaction of passengers.
Cloud modeDeep learningCustomer behaviorTravel index
国家科技期刊平台
NSTL
万方数据
