| Under the blessing of the development of the Internet,the civil aviation industry has continued to rise.At the beginning of the 21 st century,large domestic airlines began to build e-ticket systems.In 2019 before the COVID-19 epidemic,the entire industry completed 659.9342 million passengers,and domestic routes completed 585.6799 million passengers.The direct sales channels of a single large airline in China trades more than 120,000 tickets every day,and transaction volume exceeds 100 million yuan.In order to keep abreast of e-ticket,membership and other business data and strengthen the monitoring of such a large-scale e-ticket system,the airline e-commerce marketing department proposed the demand to build an e-ticket monitoring system to achieve dual monitoring of orders and fares.This thesis takes the civil aviation e-ticket platform as the project background and discusses the real-time large screen display data and monitoring requirements of the eticket monitoring system,and completes the design and implementation of the system.This thesis discusses in detail the comparison of different monitoring strategies during the system construction process.The regular monitoring function of e-tickets uses the average number of orders in the same period in the past few weeks as a reference.In such a monitoring strategy,one is that any historical data is too high or too low,the current monitoring will be inaccurate;the other is that dirty data will be generated under abnormal conditions,which requires manual processing so as not to affect subsequent monitoring,and this processing is very cumbersome.So the regular monitoring function is very passive.Data is the most important asset in the 21 st century,and artificial intelligence is in a period of rapid development.In order to solve the problems in the above monitoring functions,this thesis proposes a monitoring strategy based on machine learning.First,it is hoped that the calculation results of the machine learning algorithm can make the reference data more flexible and more in line with the trend of data changes;second,it is hoped to reduce manual interference in the monitoring function and further improve the automation level of e-ticket monitoring.In summary,this thesis aims to implement a civil aviation e-ticket monitoring system,and combine machine learning algorithms to explore a more accurate and flexible monitoring strategy. |
