Semantics for Keyword Targeting Systems

OpenAmplify (, the Semantic Web platform, today launched PinPoint™, a new service that significantly improves the effectiveness of online advertising by delivering semantic data into keyword targeting for the very first time. PinPoint can be implemented by ad networks, ad exchanges and online publishers to intelligently identify the most relevant and appropriate keyword for ad targeting on a Web page, based on sentiments, intentions and advice expressed on that page.

Although widely used throughout the industry, keyword targeting, which uses word frequency as a determinant of topic, is fundamentally unintelligent and often proves inaccurate. PinPoint, seen in action here:, is advancing the process by using semantic data and meaning that is taken from a Web page to determine more relevant keywords for advertising.

“With PinPoint we are breaking new ground. Many companies now understand the value of semantic data. There is industry-wide interest in harnessing this technology, but few companies are in a position to quickly integrate, apply the data in a meaningful way and manage it at scale. We wanted to make it easy. PinPoint is the game changer that adds the semantic layer to the keyword-oriented advertising world, bringing a new level of intelligence to ad targeting,” said Mark Redgrave, CEO of OpenAmplify.”

Ad systems that use any form of keyword targeting can use OpenAmplify’s PinPoint to improve selection of keywords. On any given Web page that mentions multiple topics, PinPoint uses the semantic context of each topic (sentiment, intent and guidance signals) to determine which topic is the most relevant.

Ad networks, ad exchanges and large publishers do not need to re-engineer their systems for PinPoint. Any system that can process a keyword can process PinPoint. PinPoint does not need taxonomy to determine better keywords, and therefore eliminates arbitrary “categorization” decisions. Even the simplest, keyword-driven ad targeting system can realise performance improvements using PinPoint on their existing infrastructure.

The impact of PinPoint on keyword targeting

As an example, consider the following text:

“People used to talk about keywords all the time. Everywhere I went it was keywords and more keywords. These days, however, traditional keywords are looking pretty limited. You should consider semantics to make things more accurate.”

Most significant topic identified by traditional keyword targeting = “keyword”
Most significant topic identified by PinPoint = “semantics”

Normal keyword targeting systems would detect the frequency of the term “keyword” and would then make a judgment that the most significant topic in the text is “keyword.” Based on semantic analysis and the meaning of the text, however, PinPoint determines that the main topic and point of interest in this text is actually “semantics.” By bringing a human understanding of the author’s true intent to the topic selection, PinPoint brings the best possible keyword to the top of the list.

Like this article?

Share on Twitter
Share on Linkdin
Share on Facebook
Share via email

Other posts that might be of interest