Research On Key Technologies Of Pre-sale Tickets Price Forecasting System

With the rapid development of aviation industry and e-commerce market,More and more aircraft are being chosen as one of the important means of transport.However,Airlines in order to obtain the maximum benefits,usually with a privatedata with the dynamic pricing strategy for pre-sale ticket pricing,for the consumer wants in volatile ticket prices find the most appropriate time to buy.For ticket itself,as a special commo dity,it is time-sensitive,high uncertainty and volatility and other characteristics,and it effect on departure time,airline,purchased in advance the number of days the comprehensive effects of multiple factors.With data mining technique is becoming more and more widely applied in the field of price forecasting,this paper will use the data mining technology to explore the association rules of the influence factors and price forecasting problem.The main content of this paper focused on the application of the data mining technology in the field of airline ticket pre-sale price analysis.From the consumer saving purchase cost point of view to carry out the following three aspects,respectively is:ticket prices influence factors data automatic acquisition and preprocessing,ticket pre-sale price and the influencing factors of association rules analysis,ticket pre-sale price forecast comparison of model and prototype system analysis and design.On the basis of existing theories and methods,the use of the web crawler technology on Qunar network take up nearly seven months of pre-sale price data as the basic data,using the data around the main research content to carry out related research.First of all,according to the characteristic of this paper data sets and the research needs to preproce ss data;And then,Based on the unclear relationship between the influence factors of the ticket pre-sale price and its booking discount,the analysis method based on classification rules,to find the relationship between the influencing factors of ticket pre-sale price and the pre-sale discount;Finally,on the premise of the influence factors of the above analysis,the multiple regression analysis,regression tree and model tree to build a predictive model to forecast ticket price and using the method of prototype system to verify.