Design And Implementation Of Online Car-hailing Itinerary Safety Guarantee System

With the rapid development of the Internet and the rush of the sharing economy,people's choice of travel methods is becoming more and more diverse.Traditional public transportation and taxis have been unable to meet people's pursuit of travel efficiency,so online car-hailing platforms emerged and quickly occupied the travel market.However,due to the non-standard industry,the personal and property safety of users in the process of using online car-hailing cannot be guaranteed,and users face various risks that are urgently needed to be solved by online car-hailing platforms.Ensuring user travel safety requires analysis and mining around itinerary data,that is,user itinerary data is the core of ensuring itinerary safety.The current platform order amount has reached an average daily level of tens of millions,the data scale continues to grow,and user itinerary data belongs to multi-source heterogeneous data.Not only is the data type complex,but also the data system is confusing.And the data cleaning processes and sample feature calculation rules are inconsistent,which make it difficult to use these data efficiently to provide safety for users ' itineray.Faced with such problems,this paper proposed an online car-hailing itinerary safety guarantee system,which aims to provide services through a unified data platform to meet the needs of the company's operators,algorithm engineers and other roles to use the itinerary data and tap its value.The system uses big data technology to process the itinerary data,as the basis of the function realization of every module.Through the feature development,accident review and data board function,it to help system users analyze and use the itinerary data.The data processing module uses Hive and Flink to process the full amount of historical data and real-time order data.The system provides a visual Web operation interface based on the B/S architecture.The MVC mode is used to separate the front and back ends to achieve decoupling.The back end is developed using frameworks such as Springboot to ensure the security and stability of the system.In order to identify sexual harassment risk orders,the XGBoost algorithm is used to train a binary classification model,so as to realize the early detection and timely intervention of sexual harassment risk orders.This paper described the complete process of system design and implementation from the aspects of system requirements analysis,summary design,detailed design and implementation and system testing,combined with use case diagrams,ER diagrams,flow charts,class diagrams,sequence diagrams,realization effect diagram and data design tables and test case tables.In the design and development of the system,I participated in the preliminary system research work and requirements analysis,and focused on the code writing of each module of the entire system,and was also responsible for the later system testing and on-line deployment.At present,the system has met various requirements and delivered to the company,and the system is in good condition.