|
Abstract:
|
In many studies, online advertising has been shown to benefit greatly from advertisement personalization. Numerous approaches focus on behavioral targeting, where advertisements are adopted to the viewer based on search term usage or web page visits. However, several authors emphasize the importance of including demographic information in web targeting practices, for brand building purposes or simply to accompany the widespread practice of defining product target groups in terms of demographic characteristics. Typically, this type of information is either hard or expensive to obtain in an online environment. In this study, a practical and cost-effective methodology, based on anonymous web usage data, is proposed for the creation of demographic web site visitor profiles that can aid web advertising targeting using Random Forests methodology.
|
