How It Works

Sophisticated algorithm—simple advantage

NetSeer’s approach to contextual analysis mirrors the way the human brain parses information to derive meaning. Our technology crawls the web extracting concepts at the page level, analyzing their interrelationships in real time. It does so by locating those concepts and their groupings on our NetSeer Concept Mapper, the first three-dimensional representation of all known concepts that form the basis of human thought.

Containing more than 50 million concepts and 2 billion relationships, NetSeer Concept Mapper has charted all the raw material of thought and the clusters into which it is grouped, to correctly identify meaning.

Understanding Context

At the keyword level, the word “bass” can mean several things – a fish, a musical instrument, a tonal range. To accurately predict user intent and good placement for your advertising you need to understand the meaning. Keyword targeting lacks the sophistication to make distinctions in meaning. Semantic analysis can assist with determining meaning, but is limited in its ability to scale. It simply cannot stand up to the demands of real-time analysis across the web. The result of this can be mismatches between your ad content and user intent, wasting your media dollars.

The Advantage for Online Advertisers and Publishers

NetSeer technology goes beyond keyword and semantic analyses. It identifies and extracts all the concepts on the page. It matches their interrelationships to the clusters of concepts located within NetSeer Concept Mapper. In this example, the concept “bass” is linked to “bass player,” “playing bass,” “acoustic bass” and “electric bass guitar” and so on, to pinpoint a cluster of concepts which identify the meaning of “bass” in this instance as a type of musical instrument. NetSeer Concept Mapper analysis eliminates the incorrect meanings related to “music” and “fish,” ensuring a relevant match with inferred user intent.

And just like the human brain, our technology makes associations beyond actual page content. To continue our example, “bass player” and “electric bass guitar” might stir an association with “Paul McCartney,” “The Beatles” or “rock and roll,” which although they might not appear in the content itself, are highly contextually relevant. Unlike keyword and semantic analyses, our concept-based approach and NetSeer Concept Mapper can extend relevancy to off-page context, going beyond the words on a page.

Matching User Intent

View some real life examples of the limitations of keyword and semantic approaches to inferring user intent.