Online advertising using linguistic knowledge

Abstract

Pay-per-click advertising is one of the most paved ways of online advertising today. However the top ranking keywords are extremely costly. Since search terms have a "long tail" behaviour, they may be used for a more cost-effective way of selecting the right keywords, achieving similar traffic, and reducing the cost considerably. This paper proposes a methodology that, exploiting linguistic knowledge, identifies cost effective bid keyword in the long tail distribution. The experiments show that these keywords are highly relevant (90% average precision) and better targeted than those suggested by other methods, while enabling reduced cost of an ad campaign. © Springer-Verlag 2011.