A practical approach to feature selection
ML92 Proceedings of the ninth international workshop on Machine learning
Computational Statistics & Data Analysis - Nonlinear methods and data mining
Predicting clicks: estimating the click-through rate for new ads
Proceedings of the 16th international conference on World Wide Web
Introduction to display advertising: a half-day tutorial
Proceedings of the fourth ACM international conference on Web search and data mining
Bid landscape forecasting in online ad exchange marketplace
Proceedings of the 17th ACM SIGKDD international conference on Knowledge discovery and data mining
Automatic ad format selection via contextual bandits
Proceedings of the 22nd ACM international conference on Conference on information & knowledge management
| Hi-index | 0.00 |
Display advertising has been growing rapidly in recent years, with revenue generated from display ads placed on spaces allocated on publisher's web pages. Traditionally, the design and layout of ad spaces on a web page are predetermined and fixed for the publisher. The objective of this work is to investigate the revenue opportunities of changing the ad layout dynamically for the publisher. A dynamic ad layout revenue optimization framework is developed for display advertising, in terms of both guaranteed and non-guaranteed advertising. The system automatically selects the ad layout template with the highest potential revenue yield for each single web page presented to the user. Forecasting algorithms are developed to predict the revenue of each ad opportunity. Two objectives are explored for the forecasting algorithms of ad layout optimization, the expected revenue and actual revenue. Promising results are obtained in offline simulation on real data collected from a Yahoo! property. The dynamic ad layout optimization system is further tested on real-time traffic and a significant revenue gain is observed compared with a static ad layout serving method.