Google Patent Hints at Ad Relevance Factors for Contextual Ads

A new patent application from Google gives hints about its plans to include relevancy factors in its contextual ad algorithms. According to Bill Slawski at SEO by the Sea, the filing describes a process of scoring ads by price, relevancy, and performance, using a technique that has popped up in several recent patent applications, "probabilistic hierarchical inferential learned cluster."

Google has been applying relevancy to ad algorithms since the fall, and causing problems for advertisers that are not factoring in Google's Quality Score when planning campaigns.