Analysis Of Railway Student Passengers’ Ticket Buying Behavior Based On Data Mining

In recent years,the length of railway operation has been increasing continuously.By the end of 2019,China’s railway mileage in operation had reached 139,000 km,an increase of 8,000 km compared to the end of 2018.As the length of railway service has increased,so as the number of passengers travelling by train.Student passenger flow is an important passenger flow of railway during spring and summer travel,which has a great influence on the number of railway passengers during spring and summer travel.Since the implementation of online ticket purchasing in 2012,the ticket system has collected a large number of information about the travel of passengers.At present,the analysis methods of railway ticket data widely used are mostly limited to the simple statistical description of the data information and lack of in-depth mining of the ticket data.According to the theoretical method and process of data mining,this paper analyzes the railway student ticket data,mining the rules of student passenger ticket purchasing behavior,and analyzing the student passenger ticket purchasing behavior.First,according to the behavior analysis principle,the paper analyzes the behavior characteristic factors of railway students’ ticket purchasing behavior,determines the purpose of data mining,and combines the theory and method of data mining to analyze the purpose of railway students’ ticket purchasing behavior,and selects the appropriate data mining method--combining statistical analysis with cluster analysis.Then,according to the process of data mining,the selected railway student ticket data is normalized.Based on the railway student ticket data from Shenyang in January 2017,according to the principle of data normalization,the original data were normalized and the results were stored in the new data warehouse.The correlation analysis method is used to determine that the main behavior characteristic factor of railway students’ ticket purchasing data is the time on the way,and the correlation between other characteristic factors that describe student passengers’ ticket purchasing behavior,such as the number of days of advance ticket purchasing,the distance on the way,seat alias,and the main characteristic factors is analyzed.Comparing the results of two-step clustering algorithm,K-means clustering algorithm and DBSCAN clustering algorithm to determine the appropriate clustering algorithm.The DBSCAN algorithm is used to perform cluster analysis on the data.Finally,the behavior rules of student passengers buying tickets are obtained from the data,and the reasons behind the behavior rules of student passengers buying tickets are analyzed based on the relevant principles of behavior analysis,and corresponding Suggestions are put forward according to the problems.