Design And Implementation Of An Online Car-hailing Advertisement System Based On ElasticSearch

With the development of the mobile Internet and the popularization of smart phones,online car-hailing has gradually become the preferred way for people to travel short distances.At the same time,the data center generates tens of millions of travel orders every day,and these orders contain great potential value.Internet advertising is an important way to tap the value of orders.Different from traditional online advertisements,online car-hailing advertisements rely on the vehicle smart screen as the carrier.After the passenger’s authorization,they can push advertisements on the car.By constructing an advertising system suitable for online car-hailing scenarios,combining travel big data analysis to achieve precise matching between advertisements and users,and at the same time promote revenue growth between online car-hailing companies and advertisers,achieving a tripartite win-win situation.We designed a commercial advertising monetization model of online car-hailing,through in-depth research on mainstream advertising platform technology and online car-hailing itinerary nodes,design a set of reasonable and usable advertising system,combined with mature Elastic Search technology,design the business logic of delivery,and combine The requirements point out the business scenarios and goals for the system,design and implement the relevant functions of the online car-hailing advertising system,and combine the TF-IDF algorithm and the BM25 algorithm to study the core data storage and search functions.The main work of the paper is as follows:(1)According to the actual business scenario requirements,analyze the advertising system,combine related development technologies,briefly describe the overall design ideas of the system,and configure platform modules,delivery function modules,distributed data storage and query module is designed and implemented in detail.(2)The first to introduce the concept of geofencing in the advertising system,improve the way of advertising.By accessing a large amount of online car-hailing order data and processing the data with the help of MQ message queues,data coupling is greatly reduced and the scalability of the system is significantly improved.(3)For big data scenarios,My SQL is used to store raw data,and Elastic Search middleware is deployed on the upper layer to achieve the required search functions.In addition,the writing speed of Elastic Search is optimized in combination with business scenarios,and the BM25 algorithm is combined on the basis of the TF-IDF algorithm to analyze in detail and design a search scoring strategy that meets the business scenario.(4)Combine the designed test cases to test the system.In the corresponding test environment,according to the designed test cases,all test points are covered,and bugs are optimized.Compared with the current exploration of commercial advertising monetization in the online car-hailing industry,the system designed and implemented in this paper has many advantages:(1)Established an integrated advertising configuration platform for internal and external(2)Introduce new delivery conditions to accurately deliver targets(3)Optimized the writing speed in combination with the scene(4)It is portable,and the delivery function can be used at any time with the newly added interface(5)High concurrency and large amount of dataThe system has been piloted in some cities.Judging from the results of a quarterly trial operation,the system has improved the efficiency of advertising management,met the needs of enterprises to develop new businesses,and achieved the strategic goal of commercial realization.