Predicting web site audience demographics using content and design cues

The size and dynamism of the Web pose challenges for all its stakeholders, which include producers and consumers of content as well as advertisers who want to place advertisements next to relevant content. A critical piece of information for producers/publishers of content as well as advertisers is the demographics of the consumers who are likely to visit a given web site. In this article we explore predictive models that attempt to deduce the demographics of the audience of a web site using cues embedded in the design or the content of its homepage. We find that it is possible to effectively predict different types of demographics of consumers of web sites on the basis of the suggested approach. Through a statistical analysis we observe that several design elements and the content differ significantly among web sites dominated by consumers of different demographic classes. We also suggest the use of an ensemble classifier that combines the content and design cues with the goal of further improving the prediction performance.