The Interactive Advertising Bureau (IAB) has released a version 2 of their taxonomy as of the first of March 2017. The new taxonomy contains more topics than the old and has gone through a general overhaul to make it more clear.
We have build a new classifier, IAB Taxonomy V2, that conforms with the latest standard.
The new ‘Content’ category has been left out but you can get content language by calling our Language Detector.
Any feedback is appreciated and we may add more training data if necessary.
Class name format
The new taxonomy has up to 4 tiers this is reflected in the class names. The format of the class names is level1_leaf_id1_id2_id3_id4 the ids correspond to the IAB codes and are integers.
You can read more about the taxonomy at their homepage where you also can find the complete id mapping.