面向流量经营敏捷推荐平台的设计与实现
发布时间：2018-11-04 07:45
【摘要】：随着移动互联网的飞速发展,运营商加速进入流量经营时代。在传统运营商管理机制中,用户数据和分析结果被分散在诸多不同的系统中,形成了“信息孤岛”,导致客户使用场景存在信息缺失。客户洞察、营销执行是以群体为粒度而不是个体,因此营销的颗粒度较粗,不能适应互联网时代的营销服务体系。面对用户上网行为的多元化,如何惠民利民,如何发展信息消费,如何控制信息安全等问题日趋扩大。客户使用流量普遍存在“不敢用、不会用、不好用”等问题,迫切需要培养客户使用习惯,剪除流量使用瓶颈,进一步提升当前支撑系统对流量内容推荐的便利性和准确性。本文从当前运营商的状况出发,为了锻造流量经营持续发展能力,构建起以智能管道(物理网络)和聚合平台(商业网络)为基础,以扩大流量规模、提升流量层次、丰富流量内涵为经营方向,以释放流量价值为目的的流量经营支撑系统。创新性地为个人客户开发并提供丰富的信息化应用,为不同行业提供具有显著社会效益的信息化解决方案,推动社会信息化进程,共享信息化发展成果。在研究和开发过程中主要进行了以下工作内容:首先,对当前主流的大数据云计算前沿技术进行融合研究,发现当前典型的移动互联网大数据应用平台较多,技术方案较混杂,性能及安全性达不到电信级的运营要求。其次,对敏捷推荐的相关技术进行研究,包括流处理技术、推荐算法等。项目基于大数据采集技术研发“基于内容指纹深度DPI识别技术”,进行应用内功能和协议的深度解析,感知获取用户的上网行为信息,位置信息,终端信息等,实现互联网新技术与运营商现有的营销支撑系统的融合;通过结合新型互联网社交网络和传统专家系统的经典算法,研发“基于社交网络模型的自适应混合协同过滤推荐算法”,实现对客户的个性化内容的精准推荐;通过采用创新技术ActiveMQ+Kafka+Spark Streaming架构的流处理技术,研发“基于消息适配的内容推荐系统”,实现利用消息适配完成传统业务消息队列和实时流引擎的高速互通,实现对推荐内容的高效推送。经过测试表明,该系统可以进行运行,实现了功能要求,达到了预期目标。本文对流处理技术的开发和产业化的应用,符合目前国内大数据市场的迫切需求和国家政策的引导扶持,对国内实时营销移动传播关键技术开发与产业化应用有一定的参考价值。
[Abstract]:With the rapid development of mobile Internet, operators are accelerating into the traffic management era. In the traditional operator management mechanism, user data and analysis results are scattered in many different systems, forming "information island", which leads to the absence of information in customer usage scenarios. Customer insight, marketing implementation is based on group granularity rather than individual, so marketing granularity is relatively coarse, can not adapt to the Internet era of marketing service system. Facing the diversification of users' online behavior, the problems such as how to benefit the people, how to develop information consumption and how to control information security are becoming more and more serious. There are some problems such as "not using, not using", and so on. Therefore, it is urgent to cultivate customer's usage habits, cut off the bottleneck of traffic usage, and further improve the convenience and accuracy of the current support system for traffic content recommendation. In order to forge the capacity of continuous development of traffic management, this paper, based on intelligent pipeline (physical network) and aggregation platform (commercial network), aims to expand the traffic scale and enhance the traffic level. The flow management support system with rich traffic connotation and the purpose of releasing the flow value is the management direction. Innovative development for individual customers and provide rich information applications, for different industries to provide significant social benefits of information solutions, to promote the process of information society, to share the fruits of information development. In the process of research and development, the main contents are as follows: first, the current mainstream of big data cloud computing frontier technology fusion research, found that the current typical mobile Internet big data application platforms, technical solutions are more mixed, Performance and security do not meet telecom-grade operational requirements. Secondly, the related technologies of agile recommendation are studied, including stream processing technology, recommendation algorithm and so on. The project is based on big data acquisition technology to develop the "content based fingerprint depth DPI identification technology" to analyze the functions and protocols within the application, and to perceive and obtain the user's online behavior information, location information, terminal information, etc. Realizing the integration of the new Internet technology and the existing marketing support system of the operators; By combining the classical algorithms of the new Internet social network and the traditional expert system, we develop an adaptive hybrid collaborative filtering recommendation algorithm based on the social network model to realize the accurate recommendation of the personalized content of the customer. By using the stream processing technology based on ActiveMQ Kafka Spark Streaming architecture, the "content recommendation system based on message adaptation" is developed to realize the high-speed interworking between traditional message queue and real-time flow engine by using message adaptation. To achieve the recommendation of the content of the efficient push. The test results show that the system can run, achieve the functional requirements and achieve the desired goal. In this paper, the development of convection processing technology and the application of industrialization are in line with the urgent needs of the domestic market of big data and the guidance and support of the national policy. It has certain reference value for the development and industrialization application of the key technology of real-time marketing mobile communication in China.
【学位授予单位】：成都理工大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TP391.3
本文编号：2309159
[Abstract]:With the rapid development of mobile Internet, operators are accelerating into the traffic management era. In the traditional operator management mechanism, user data and analysis results are scattered in many different systems, forming "information island", which leads to the absence of information in customer usage scenarios. Customer insight, marketing implementation is based on group granularity rather than individual, so marketing granularity is relatively coarse, can not adapt to the Internet era of marketing service system. Facing the diversification of users' online behavior, the problems such as how to benefit the people, how to develop information consumption and how to control information security are becoming more and more serious. There are some problems such as "not using, not using", and so on. Therefore, it is urgent to cultivate customer's usage habits, cut off the bottleneck of traffic usage, and further improve the convenience and accuracy of the current support system for traffic content recommendation. In order to forge the capacity of continuous development of traffic management, this paper, based on intelligent pipeline (physical network) and aggregation platform (commercial network), aims to expand the traffic scale and enhance the traffic level. The flow management support system with rich traffic connotation and the purpose of releasing the flow value is the management direction. Innovative development for individual customers and provide rich information applications, for different industries to provide significant social benefits of information solutions, to promote the process of information society, to share the fruits of information development. In the process of research and development, the main contents are as follows: first, the current mainstream of big data cloud computing frontier technology fusion research, found that the current typical mobile Internet big data application platforms, technical solutions are more mixed, Performance and security do not meet telecom-grade operational requirements. Secondly, the related technologies of agile recommendation are studied, including stream processing technology, recommendation algorithm and so on. The project is based on big data acquisition technology to develop the "content based fingerprint depth DPI identification technology" to analyze the functions and protocols within the application, and to perceive and obtain the user's online behavior information, location information, terminal information, etc. Realizing the integration of the new Internet technology and the existing marketing support system of the operators; By combining the classical algorithms of the new Internet social network and the traditional expert system, we develop an adaptive hybrid collaborative filtering recommendation algorithm based on the social network model to realize the accurate recommendation of the personalized content of the customer. By using the stream processing technology based on ActiveMQ Kafka Spark Streaming architecture, the "content recommendation system based on message adaptation" is developed to realize the high-speed interworking between traditional message queue and real-time flow engine by using message adaptation. To achieve the recommendation of the content of the efficient push. The test results show that the system can run, achieve the functional requirements and achieve the desired goal. In this paper, the development of convection processing technology and the application of industrialization are in line with the urgent needs of the domestic market of big data and the guidance and support of the national policy. It has certain reference value for the development and industrialization application of the key technology of real-time marketing mobile communication in China.
【学位授予单位】：成都理工大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TP391.3
【参考文献】
中国期刊全文数据库 前1条
1 杨霞;吴东伟;;R语言在大数据处理中的应用[J];科技资讯;2013年23期
，本文编号：2309159
本文链接：https://www.wllwen.com/guanlilunwen/yingxiaoguanlilunwen/2309159.html
