在线广告中点击率预测研究

发布时间：2018-02-28 13:17

本文关键词： 计算广告 CTR 机器学习　出处：《华东师范大学学报(自然科学版)》2017年05期 　论文类型：期刊论文

【摘要】：随着互联网的发展和用户的增长,广告行业从传统的线下广告模式,逐步转变为线上广告模式.同时,由于大数据分析技术的运用,线上广告模式相比于传统广告也体现了巨大的优越性.广告主之间相互竞争,通过竞价的方式,将自己的广告投放在运营媒体的广告位上.所以,在投放前预测该广告可能被用户点击的概率(CTR),对于广告主减少成本和增加可能收入来说非常重要.本文在调研了目前常用的广告点击率预测模型的基础上,选取广告主、广告和投放媒体平台信息作为预测模型的特征,采用真实数据集验证说明各种模型的优劣性,以及不同特征对广告点击率预测结果的影响.
[Abstract]:With the development of the Internet and the growth of users, the advertising industry has gradually changed from the traditional offline advertising model to the online advertising mode. At the same time, due to the use of big data's analytical technology, Compared with traditional advertising, online advertising mode also shows great superiority. Advertisers compete with each other and place their advertisements in the advertising positions of the operating media by way of bidding. It is very important for advertisers to reduce the cost and increase the possible revenue by predicting the probability that the advertisement may be clicked by the user before putting it into service. This paper selects the advertiser on the basis of investigating the commonly used ad click rate prediction model. Advertising and media platform information is used as the feature of the prediction model. The real data set is used to verify the advantages and disadvantages of the models and the influence of different features on the prediction results of ad click rate.
【作者单位】：华东师范大学数据科学与工程学院;长江口水文水资源勘测局;
【基金】：国家重点研发计划(2016YFB1000905) 国家自然科学基金广东省联合重点项目(U1401256);国家自然科学基金(61672234,61402177) 华东师范大学信息化软课题
【分类号】：TP181;TP393.092

【相似文献】