X
Lotame Named DMP of the Year! Selected for Superior Technology & Independence Learn More

Lotame’s Semantic Classification Technology

July 18, 2016

Lotame’s Semantic Classifier is a Natural Language Processing (NLP) engine that identifies the concepts and context of an article. Semantic Classification relies on article content to capture user interests at a far more granular scale when compared to standard page-level data collection methods. Lotame’s semantic classification technology uses the actual content of an article page to identify user interests and categorize that user into pre-defined behavioral nodes. (For more basic information about Semantic Classification, check out our Back to Basics: What is Semantic Classification article.)

Why Use Semantic Classification?

Lotame’s Semantic Classification solves a few large problems facing publishers today:

Using standard data collection techniques, most publishers are restricted by the metadata, breadcrumbs, or URL structure of their content to classify articles and user interests. By leveraging Lotame’s Semantic Classification engine, publishers can now unlock the power of their content to identify previously unknown user interests.

What Data is Gathered with Semantic Classification?

Semantic Classification delivers insight into user interests at a far more granular level, generally not available using article metadata. Publishers using Lotame now have thousands of concepts at their fingertips to classify user behavior. Those hard-to-find, highly sought-after audiences command higher CPMs and satisfy discerning advertisers that are looking to find users beyond the “typical” audience buckets.

Why settle for selling a “Cooking” or “Entertainment” audience at a relatively low CPM when you can deliver tailor-made audiences, like “Baking Holiday Sweets” or “Live Jazz” for a significantly higher CPM? Your advertisers will thank you by returning again and again for the audiences they’re unable to find anywhere else.

We love editors, but…

Do you question the accuracy of your article metadata? Are your articles frequently cross-posted in different site sections to increase pageviews and traffic? That’s a standard practice across the industry, but it may render the URL structure of your articles useless for classifying content. By leveraging Lotame’s Semantic Classification technology, you can stop relying on metadata appended by editors (or not) and start using the actual concepts pulled directly from the content itself.

Semantic Classification also works in conjunction with metadata / breadcrumbs, as well as in situations where metadata / breadcrumbs are insufficient.

Keep in Mind: 

  • This may be obvious, but language processing relies on language available on a particular webpage. We aren’t able to classify photos, videos, or other non-text content.
  • Similarly, we are experts at classifying English language content. If there are other languages you would like us to learn, please contact us.
  • We understand the need to classify the sentiment of an article, but right now we can only tell you what an article is about, not whether the tone of the article is positive or negative.

Once Semantic Classification is enabled for your account, data will begin flowing into the Standard Hierarchy in approximately 24-48 hours.

For more information, connect with your Client Success Manager (CSM) or contact Lotame to see how Lotame’s Semantic Classification technology can develop your audience targeting abilities.