智能广告投放平台系统的设计与实现

发布时间：2018-06-12 02:06

本文选题：广告投放系统 + Java　；参考：《北京交通大学》2017年硕士论文

【摘要】：如今,广告投放主要可分为两种形式。第一种是线下广告,如海报、广告机轮播图片、屏幕播放视频广告等。另一种是线上广告,如各大网站通过广告平台投放的广告。各种广告投放方式都有各自的优缺点,尤其是线下广告。线下广告一般是一种张贴布置的形式,位置十分灵活,但是缺点是显而易见的,没有目的性,互动性差,不能吸引人的注意。智能广告投放平台系统正是为了解决以上问题而出现的。系统中通过人脸识别技术获取人群数据,根据不同人的属性数据定向投放广告,并且返回广告投放效果,达到更为精准的投放广告的目的。系统主要包括广告投放业务管理系统、广告机监控模块、接口服务模块、广告机前端模块、大数据分析平台五个部分。笔者在项目开发中,参与完成系统整体需求分析、架构设计、详细设计和实现的全过程。独立完成广告投放业务管理系统中四个子模块和广告机监控模块的架构设计和实现。五个模块功能如下:(1)统计分析模块:该模块为系统管理员提供访客统计功能,为商家和广告主提供投放效果分析、单条广告效果分析功能。(2)账户管理模块:为系统管理员提供对广告主和商家的账号管理功能。包括查看账号信息、开通账号户、停用账户、修改账户信息。(3)投放管理模块:该模块为用户提供对投放到平台的广告进行管理的功能。其中包括获取平台中所有广告列表、新建广告、修改广告。(4)计费管理模块:该模块为用户提供费用查询管理功能。其中包括返回商家分成信息、获取分成信息明细、获取广告计费信息、获取单条广告费用明细、确认收款五个功能。(5)广告机监控模块:该模块实现远程地对分布在各个不同位置的平台中广告机的管理。其中包括查询广告机的信息,修改广告机的配置信息。系统使用Spring、SpringMVC、Mybatis框架为基础。前端Web界面,使用VUE框架开发。业务数据存储在MySQL数据库中,Redis保存缓存数据。图片储存在分布式图片存储系统FastDFS中。Kafka保存日志数据。基于Netty框架、ZooKeeper和RabbitMQ消息队列,搭建分布式集群监测服务模块,用于保证系统的可靠性,通过心跳方式监控广告机在线状态,修改广告机的一些配置信息。
[Abstract]:Today, advertising can be mainly divided into two forms. The first is offline advertising, such as posters, advertising machines on the broadcast of pictures, and video ads on screen. The other is online advertising, such as advertising on the advertising platform. All kinds of advertising have their own advantages and disadvantages, especially under line advertising. The position of a posting arrangement is very flexible, but the shortcomings are obvious, there is no purpose, poor interaction, and can not attract people's attention. The system of intelligent advertising platform is just to solve the above problems. In the system, human data are obtained by face recognition technology, and the target is directed according to the attribute data of different people. Advertising, and return to the effect of advertising, to achieve more accurate advertising purposes. The system mainly includes advertising management system, advertising machine monitoring module, interface service module, advertising machine front end module, large data analysis platform five parts. In the project development, the participation of the overall system needs analysis, architecture The whole process of detailed design and implementation. The architecture design and implementation of four sub modules and advertising machine monitoring modules are completed independently. The functions of the five modules are as follows: (1) statistical analysis module: This module provides visitor statistics function for the system administrator and provides the analysis of the effect for the merchants and advertisers. Advertising effect analysis function. (2) account management module: to provide system administrators with the advertisers and merchants account management functions. Including checking account information, opening account accounts, disusing accounts, modifying account information. (3) launch management module: This module provides users with the management of advertising on the platform. Take all the advertising lists, new ads, and modify advertisements. (4) charge management module: This module provides the user with the cost query management function. It includes returning the business into information, obtaining information details, obtaining the advertising billing information, obtaining the details of the single advertising cost, confirming the five functions of the receipts. (5) the advertisement machine monitoring module: (5) the advertisement machine monitoring module: The module implements the management of advertising machines distributed in different locations remotely, including the information of the query advertising machine and the modification of the configuration information of the ad machine. The system is based on the Spring, SpringMVC, Mybatis framework. The front-end Web interface is developed with the VUE framework. The business data is stored in the MySQL database, and the Redis is stored in the cache number. According to the data stored in the distributed picture storage system FastDFS,.Kafka saves log data. Based on the Netty framework, ZooKeeper and RabbitMQ message queues, a distributed cluster monitoring service module is built to ensure the reliability of the system, monitor the online state of the ad machine by heartbeat, and modify some configuration information of the ad machine.
【学位授予单位】：北京交通大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP311.52

【参考文献】