Robust Planning of Integrated Targeted Display Advertising
Researcher(s)
Description
Online advertising publishers typically sell their ads via spot market and upfront market. In an upfront market, ad resources are sold via contracts whereby an ad publisher promises to deliver ce1iain amount of resources for advertisers over a period of time, so-called guaranteed ads. In a spot market, ad publishers sell their ad resources through simple auction mechanisms. It is a challenging problem for ad publishers to maximize their spot market revenue in the meanwhile of fulfilling their guarantees.In this project, we propose to address the ad planning problem with a robust approach; specifically, we will propose a distributionaly robust chance-constrained stochastic programming model to allocate in real time the amount of resources for spot markets and for upfront market. We will develop efficient algorithms for solving the large-scale instances in practice. Completion of the project will greatly boost the ad resource management performance of ad publishers.
Detail(s)
| Project number | 7004269 |
| Status | Finished |
| Effective start/end date | 1/10/14 -> 7/04/16 |