Design And Implementation Of Ticket Analysis System Based On Hadoop

In recent years,China attaches great importance to the status and role of big data for social and economic development,and sufficient policy support has made the technology and application of big data develop significantly.At the same time,online ticketing benefits from the popularity of mobile Internet and the stable growth of business scale.It carries out secondary development analysis on the massive data generated,provides data support for decision-making,stimulates business model innovation,spawns new business growth points,and finally occupies more market share,becoming an important driving force for major Internet companies.Based on the actual project of a company,this paper designs and implements a ticket analysis system based on Hadoop.The system provides the whole process analysis ability for each business line of online ticketing,and effectively supports daily operation and decision-making.According to the idea of software engineering,this paper first describes the background and significance of the ticketing analysis system.According to the multi-dimensional business,heterogeneous massive data and different timeliness requirements faced by the analysis work,it gives a technical solution based on Hadoop ecology.Secondly,combined with the role and use case diagram,the functional and non functional requirements of the system are given,including the data collection function of user behavior log and ticketing business data,the real-time and offline analysis function of home cockpit,movie tickets,performance tickets,advertising and member marketing,the result visualization display function and user management function.Then,through the overall design,the system architecture is determined by layers,which is divided into data storage layer,data collection layer,data analysis layer and web service layer from bottom to top.The data storage layer uses HDFS,MySQL and redis to store and dredge the system data.The data collection layer includes user behavior log data collection and ticket business MySQL data collection.The data analysis layer uses spark for real-time and offline calculation of ticket indicators.The web service layer is used as the system access interface to visualize the results and provide the ability of user management.Aiming at the multi-dimensional business of ticketing,a five tier data warehouse is designed to improve the data governance ability.Finally,the ticketing analysis system is designed in detail,and the functional sequence diagram,flow chart and data model relationship are determined and coded to complete the construction of the ticketing analysis system.This paper finally completes all the work of the ticketing analysis system,its independent operation provides the business personnel with the whole process of business data management and intuitive multi-dimensional analysis ability,effectively improves the work efficiency,and is the effective guarantee for the implementation of ticketing business and the construction of large cultural and entertainment system.