“Fashion blog.” “Political news.” “Drone enthusiast forum.”
It’s not hard to come up with a broad categorization for any page online. Most 3rd-party data technologies assign those categorizations subjectively by humans for each page event for each user, building up user profiles that get bundled into target segments.What’s missing here? Take the fashion blog example: what if there are posts about men’s fashion, and also about women’s? About jeans? Dresses? Hats? “Fashion blog” is such a broad category that a fashion retailer looking to target consumers would struggle to find a specific audience segment. This lack of information – lost in the processing of page visit events – is what we call information destruction.
There’s a better way to process page events, one that isn’t prone to human biases, and that preserves information nuances on each page. Using the machine learning discipline natural language processing (NLP), Semasio has unbiased artificial intelligence “read” the page and glean important information.
Intelligently programmed to differentiate between editorial content and everything else, NLP pulls the most important words, phrases and sentiment from a page. So each page on a fashion blog is processed according to the specific subject, not just by a general category.
All target segments start with user profiles, and all user profiles start with an individual event. By preserving all of the valuable information from these events, Semasio creates far more nuanced and defined target segments.
What’s more, our NLP AI is constantly learning – every new event adds to its vocabulary and understanding. The segments created dynamically evolve, meaning things never freeze in time or go stale, keeping pace with your audience's unique evolving interests and intents.