网络定向广告投放系统研究
发布时间：2019-05-23 14:32
【摘要】：随着互联网的飞速发展,利用网络投放定向广告成为比较高效的广告投放方式。在现有的网络定向广告算法中存在着内容匹配精度较差、用户偏好考虑不周全等问题,为了解决这些问题,本文提出了基于语义簇的内容定向广告算法,并在此基础之上,考虑用户的交互历史,学习用户偏好,提出了一种基于用户偏好的行为定向广告算法。通过研究文献和实验验证,设计了一种网络定向广告投放系统。本文主要进行了以下方面的工作:1.为了提升广告和网页内容的匹配精度,本文提出了一种基于语义簇内容定向广告算法。在算法中定义了语义簇的概念,并基于此改进了传统的TF-IDF词权重度量方法。经关键字抽取和基于词义相关度的语义扩展后,计算网页和广告文本的相似度。设计实验对算法的效果进行了验证。2.为了使广告投放更加精准,在内容定向的基础上加入了用户偏好。根据用户的日常交互信息学习用户的个人偏好。根据事先得到的内容匹配的待投放广告,计算基于用户偏好的广告权重,为用户投放权重最高的广告。设计实验对算法的效果进行了验证。3.对网络定向广告系统进行了整体设计,包括用户交互模块、数据收集与处理模块,广告定向投放模块和数据存储模块。经过实验验证,本文所设计的网络定向广告系统,在投放精度上,达到了预期效果。
[Abstract]:With the rapid development of the Internet, the use of network targeted advertising has become a more efficient way of advertising. In order to solve these problems, there are some problems in the existing network oriented advertising algorithms, such as poor content matching accuracy and incomplete user preference. In order to solve these problems, this paper proposes a content oriented advertising algorithm based on semantic cluster, and on this basis, Considering the interaction history of users and learning user preferences, a behavior oriented advertising algorithm based on user preferences is proposed. Through the research literature and experimental verification, a network oriented advertising system is designed. The main work of this paper is as follows: 1. In order to improve the matching accuracy of advertising and web content, this paper proposes a semantic cluster content oriented advertising algorithm. The concept of semantic cluster is defined in the algorithm, and based on this, the traditional TF-IDF word weight measurement method is improved. After keyword extraction and semantic extension based on word meaning relevance, the similarity between web page and advertising text is calculated. The effect of the algorithm is verified by the design experiment. 2. In order to make advertising more accurate, user preference is added on the basis of content orientation. Learn the user's personal preferences according to the user's daily interactive information. According to the content matching advertisement to be put in advance, the advertising weight based on user preference is calculated, and the advertisement with the highest weight is given to the user. The effect of the algorithm is verified by the design experiment. The network oriented advertising system is designed as a whole, including user interaction module, data collection and processing module, advertisement oriented delivery module and data storage module. The experimental results show that the network oriented advertising system designed in this paper has achieved the desired results in terms of delivery accuracy.
【学位授予单位】：北京邮电大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP393.0
本文编号：2483980
[Abstract]:With the rapid development of the Internet, the use of network targeted advertising has become a more efficient way of advertising. In order to solve these problems, there are some problems in the existing network oriented advertising algorithms, such as poor content matching accuracy and incomplete user preference. In order to solve these problems, this paper proposes a content oriented advertising algorithm based on semantic cluster, and on this basis, Considering the interaction history of users and learning user preferences, a behavior oriented advertising algorithm based on user preferences is proposed. Through the research literature and experimental verification, a network oriented advertising system is designed. The main work of this paper is as follows: 1. In order to improve the matching accuracy of advertising and web content, this paper proposes a semantic cluster content oriented advertising algorithm. The concept of semantic cluster is defined in the algorithm, and based on this, the traditional TF-IDF word weight measurement method is improved. After keyword extraction and semantic extension based on word meaning relevance, the similarity between web page and advertising text is calculated. The effect of the algorithm is verified by the design experiment. 2. In order to make advertising more accurate, user preference is added on the basis of content orientation. Learn the user's personal preferences according to the user's daily interactive information. According to the content matching advertisement to be put in advance, the advertising weight based on user preference is calculated, and the advertisement with the highest weight is given to the user. The effect of the algorithm is verified by the design experiment. The network oriented advertising system is designed as a whole, including user interaction module, data collection and processing module, advertisement oriented delivery module and data storage module. The experimental results show that the network oriented advertising system designed in this paper has achieved the desired results in terms of delivery accuracy.
【学位授予单位】：北京邮电大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP393.0
【参考文献】
相关期刊论文 前1条
1 卫强;阮楠;单艺;;网络展示广告位置对点击率CTR影响的实证研究[J];信息系统学报;2010年01期
相关硕士学位论文 前2条
1 施灿灿;网络定向广告中用户兴趣模型研究与应用[D];合肥工业大学;2013年
2 俞淑平;网络定向广告投放算法研究[D];浙江大学;2010年
，本文编号：2483980
本文链接：https://www.wllwen.com/guanlilunwen/ydhl/2483980.html
