面向流量經(jīng)營(yíng)敏捷推薦平臺(tái)的設(shè)計(jì)與實(shí)現(xiàn)
發(fā)布時(shí)間：2018-11-04 07:45
【摘要】：隨著移動(dòng)互聯(lián)網(wǎng)的飛速發(fā)展,運(yùn)營(yíng)商加速進(jìn)入流量經(jīng)營(yíng)時(shí)代。在傳統(tǒng)運(yùn)營(yíng)商管理機(jī)制中,用戶數(shù)據(jù)和分析結(jié)果被分散在諸多不同的系統(tǒng)中,形成了“信息孤島”,導(dǎo)致客戶使用場(chǎng)景存在信息缺失�？蛻舳床�、營(yíng)銷執(zhí)行是以群體為粒度而不是個(gè)體,因此營(yíng)銷的顆粒度較粗,不能適應(yīng)互聯(lián)網(wǎng)時(shí)代的營(yíng)銷服務(wù)體系。面對(duì)用戶上網(wǎng)行為的多元化,如何惠民利民,如何發(fā)展信息消費(fèi),如何控制信息安全等問(wèn)題日趨擴(kuò)大�？蛻羰褂昧髁科毡榇嬖凇安桓矣谩⒉粫�(huì)用、不好用”等問(wèn)題,迫切需要培養(yǎng)客戶使用習(xí)慣,剪除流量使用瓶頸,進(jìn)一步提升當(dāng)前支撐系統(tǒng)對(duì)流量?jī)?nèi)容推薦的便利性和準(zhǔn)確性。本文從當(dāng)前運(yùn)營(yíng)商的狀況出發(fā),為了鍛造流量經(jīng)營(yíng)持續(xù)發(fā)展能力,構(gòu)建起以智能管道(物理網(wǎng)絡(luò))和聚合平臺(tái)(商業(yè)網(wǎng)絡(luò))為基礎(chǔ),以擴(kuò)大流量規(guī)模、提升流量層次、豐富流量?jī)?nèi)涵為經(jīng)營(yíng)方向,以釋放流量?jī)r(jià)值為目的的流量經(jīng)營(yíng)支撐系統(tǒng)。創(chuàng)新性地為個(gè)人客戶開(kāi)發(fā)并提供豐富的信息化應(yīng)用,為不同行業(yè)提供具有顯著社會(huì)效益的信息化解決方案,推動(dòng)社會(huì)信息化進(jìn)程,共享信息化發(fā)展成果。在研究和開(kāi)發(fā)過(guò)程中主要進(jìn)行了以下工作內(nèi)容:首先,對(duì)當(dāng)前主流的大數(shù)據(jù)云計(jì)算前沿技術(shù)進(jìn)行融合研究,發(fā)現(xiàn)當(dāng)前典型的移動(dòng)互聯(lián)網(wǎng)大數(shù)據(jù)應(yīng)用平臺(tái)較多,技術(shù)方案較混雜,性能及安全性達(dá)不到電信級(jí)的運(yùn)營(yíng)要求。其次,對(duì)敏捷推薦的相關(guān)技術(shù)進(jìn)行研究,包括流處理技術(shù)、推薦算法等。項(xiàng)目基于大數(shù)據(jù)采集技術(shù)研發(fā)“基于內(nèi)容指紋深度DPI識(shí)別技術(shù)”,進(jìn)行應(yīng)用內(nèi)功能和協(xié)議的深度解析,感知獲取用戶的上網(wǎng)行為信息,位置信息,終端信息等,實(shí)現(xiàn)互聯(lián)網(wǎng)新技術(shù)與運(yùn)營(yíng)商現(xiàn)有的營(yíng)銷支撐系統(tǒng)的融合;通過(guò)結(jié)合新型互聯(lián)網(wǎng)社交網(wǎng)絡(luò)和傳統(tǒng)專家系統(tǒng)的經(jīng)典算法,研發(fā)“基于社交網(wǎng)絡(luò)模型的自適應(yīng)混合協(xié)同過(guò)濾推薦算法”,實(shí)現(xiàn)對(duì)客戶的個(gè)性化內(nèi)容的精準(zhǔn)推薦;通過(guò)采用創(chuàng)新技術(shù)ActiveMQ+Kafka+Spark Streaming架構(gòu)的流處理技術(shù),研發(fā)“基于消息適配的內(nèi)容推薦系統(tǒng)”,實(shí)現(xiàn)利用消息適配完成傳統(tǒng)業(yè)務(wù)消息隊(duì)列和實(shí)時(shí)流引擎的高速互通,實(shí)現(xiàn)對(duì)推薦內(nèi)容的高效推送。經(jīng)過(guò)測(cè)試表明,該系統(tǒng)可以進(jìn)行運(yùn)行,實(shí)現(xiàn)了功能要求,達(dá)到了預(yù)期目標(biāo)。本文對(duì)流處理技術(shù)的開(kāi)發(fā)和產(chǎn)業(yè)化的應(yīng)用,符合目前國(guó)內(nèi)大數(shù)據(jù)市場(chǎng)的迫切需求和國(guó)家政策的引導(dǎo)扶持,對(duì)國(guó)內(nèi)實(shí)時(shí)營(yíng)銷移動(dòng)傳播關(guān)鍵技術(shù)開(kāi)發(fā)與產(chǎn)業(yè)化應(yīng)用有一定的參考價(jià)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.
【學(xué)位授予單位】：成都理工大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2015
【分類號(hào)】：TP391.3
本文編號(hào)：2309160
[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.
【學(xué)位授予單位】：成都理工大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2015
【分類號(hào)】：TP391.3
【參考文獻(xiàn)】
相關(guān)期刊論文 前1條
1 楊霞;吳東偉;;R語(yǔ)言在大數(shù)據(jù)處理中的應(yīng)用[J];科技資訊;2013年23期
，本文編號(hào)：2309160
本文鏈接：http://www.lk138.cn/guanlilunwen/yingxiaoguanlilunwen/2309160.html
最近更新
教材專著
