Here is how our proprietary AI-powered framework works and how it allows us to offer a unique way to gauge media coverage
Discoverage uses a proprietary framework that with a combination of DOM traversing, HTML parsing and server-side rendered code analysis can understand the structure of any media homepage in any language. Our custom framework is designed to accurately extract what we call article boxes. Boxes are framed content that represent an article and can include just a text link or even a title, an image, an icon and much more.
Grading article boxes is essential to weigh coverage. The bigger and the closer to the top of the page the box is, the better is the grade. To grade articles, Discoverage uses an AI to understand how a homepage is composed and where the most important article boxes are. Then a clustering algorithm is applied to group boxes together and to grade them from AAA (the most important piece of news ) down to Z (the articles buried in the secondary part of the page)
Space and size are only two of the variables used to weigh coverage, with time being the third and last one. Discoverage can detect how articles "move" in a homepage over time to accurately calculate how long each box is present on a homepage and at which grade.Get in Touch
Finally, to gauge media coverage it's essential that we follow stories and trends. Discoverage uses a custom implementation of market standard AI frameworks to group together articles belonging to the same story, either directly or tangentially, weighing them differently of course, and to exclude any article that refers to the same people and events but does not belong.