Keyword Optimization in Sponsored Search via Feature Selection


Svetlana Kiritchenko, Mikhail Jiline ;
Proceedings of the Workshop on New Challenges for Feature Selection in Data Mining and Knowledge Discovery at ECML/PKDD 2008, PMLR 4:122-134, 2008.


Sponsored search is a new application domain for the feature selection area of research. When a user searches for products or services using the Internet, most of the major search engines would return two sets of results: regular web pages and paid advertisements. An advertising company provides a set of keywords associated with an ad. If one of these keywords is present in a user’s query, the ad is displayed, but the company is charged only if the user actually clicks on the ad. Ultimately, a company would like to advertise on the most effective keywords to attract only prospective customers. A set of keywords can be optimized based on historic performance. We propose to optimize advertising keywords with feature selection techniques applied to the set of all possible word combinations comprising past users’ queries. Unlike previous work in this area, our approach not only recognizes the most profitable keywords, but also discovers more specific combinations of keywords and other relevant words.

Related Material