WebSelect : A Research Prototype for Optimizing Ad Exposures based on Network Structure

Abstract

This paper describes a Research Prototype, WebSelect, designed to assist online media planners in deciding which websites to target for a media campaign. The salient feature of the prototype is its ability to capture and utilize the overlap information in website traffic and use it for media planning exercises. In addition, the prototype possesses the capability to include targeting parameters like users age group and income along with the varying advertising costs across different websites. The prototype uses a genetic algorithm at the backend to select the final subset to target from the possible website set.