首页> 外文会议>International Conference on Sustainable Energy Information Technology >Ad-centric Model Discovery for Predicting Ads' Click-through Rate
【24h】

Ad-centric Model Discovery for Predicting Ads' Click-through Rate

机译:以广告为中心的模型发现,用于预测广告的点击率

获取原文

摘要

Click here and insert your abstract text. Search engine advertising has become one of the most important revenue models of electronic commerce. It strongly affects the probability that users click on the ads at the side of the search results page if the system shows the right ones. To maximize the outcome of search engine revenue and improve users' perception on those ads, it is important to understand the factors which affect the click-through rate (CTR) on those ads. Tencent founded in 1998, is one of China's largest and most used Internet service portals. It provides a number of online services such as value-added Internet, mobile and telecom services and online advertising. As of September 30, 2011, Tencent had 711.7 million active Instant Messenger users. It forms the largest Internet Community in China. In this research, we use a very large dataset of Tencent click logs (soso.com) with millions records. First we describe how soso.com searching engine advertising works, our system architecture is designed with the click log dataset, and observations inside it aims at those ads with enough historical click logs. Then we show how to use ad-centric features to discover models that can find factors affecting CTR prediction performance. The proposed framework could help both optimizing the search engine system for soso.com and improving the ads designs for the advertisers.
机译:单击此处并插入抽象文本。搜索引擎广告已成为电子商务最重要的收入模型之一。如果系统显示正确的,它强烈影响用户在搜索结果页面侧面点击广告的概率。为了最大限度地提高搜索引擎收入并提高用户对这些广告的看法,重要的是要了解影响这些广告上的点击率(CTR)的因素。腾讯成立于1998年,是中国最大,最常用的互联网服务门户网站之一。它提供了许多在线服务,如增值互联网,移动和电信服务和在线广告。截至2011年9月30日,腾讯有71170万活跃的即时通信用户。它形成了中国最大的互联网社区。在这项研究中,我们使用数百万纪录的腾讯单击日志(SOSO.com)的非常大的数据集。首先,我们描述了SOSO.com搜索引擎广告工作的方式,我们的系统架构采用点击日志数据集设计,并且它内部的观察结果旨在拥有足够的历史点击日志的广告。然后,我们展示了如何使用广告中心来发现可以找到影响CTR预测性能的因素的型号。建议的框架可以帮助优化SOSO.com的搜索引擎系统,并改善广告商的广告设计。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号