I came across an article on the New Scientist website about the latest development in search engines, and how some interesting new approaches are being investigated. One such approach by Ingmar Weber and Carlos Castillo uses demographic data - age, gender, ethnicitiy etc, to identify how the search behavious of different groups of people differe. For example, if women put into a search engine 'Opal' they are more likely to be searching for jewellery, whereas when men use the same search term, they are more likely to be looking for information about a car.

Other possible applications for demographically based content is targeted advertising. If you type in the words 'familly holiday' then the adverts could be tailored to your income level, so high income visitors are shown expensive holidays, and low income are shown budget holidays.

The research was based on 3 months log data from Yahoo search, linking general demographic data with search intention. This was then checked over a 12 month period using search data from Yahoo to see if demographic modifiers affect the accuracy of the search engine to bring up the desired information. The result of which showed a 7% improvement. Not much, but in the world of advertising, 7% improvement in click through can add up to some very large increases in revenue.