Abstract
In online advertising campaigns, to measure purchase propensity, click-through rate (CTR), defined as a ratio of number of clicks to number of impressions, is one of the most informative metrics used in business activities such as performance evaluation and budget planning. No matter what channel an ad goes through (display ads, sponsored search or contextual advertising), CTR estimation for rare events is essential but challenging, often incurring with huge variance, due to the sparsity in data. In this chapter, to alleviate this sparsity, we develop models and methods to smoothen CTR estimation by taking advantage of the natural data hierarchy or by clustering and data continuity in time to leverage information from data close to the events of interest. In a contextual advertising system running at Yahoo!, we demonstrate that our methods lead to significantly more accurate estimation of CTRs.
Reference7 articles.
1. Agarwal, D., Broder, A. Z., Chakrabarti, D., Diklic, D., Josifovski, V., & Sayyadian, M. (2007). Estimating rates of rare events at multiple resolutions. In Proceedings of the 13th ACM SIGKDD international Conference on Knowledge Discovery and Data Mining (pp. 16-25).
2. Agarwal, D., Chen, B., & Elango, P. (2009). Spatio-temporal models for estimating click-through rate. In Proceedings of the 18th international Conference on World Wide Web (pp. 21-30).
3. Algorithm AS 103: Psi (Digamma) Function
4. Interactive Advertising Bureau (IAB) & PricewaterhouseCoopers International. (PwC). (2009). Internet ad revenues at $10.9 billion for first half of ’09. Retrieved December 15, 2009, from http://www.iab.net/about_the_iab/recent_press_releases/press_release_archive/press_release/pr-100509
5. On Information and Sufficiency
Cited by
17 articles.
订阅此论文施引文献
订阅此论文施引文献，注册后可以免费订阅5篇论文的施引文献，订阅后可以查看论文全部施引文献