Information annotation is a method where all the unlabelled and all the unstructured information and data are typical labeled, and they are all kept separate for just one reason that is perfect for training machine discovering models.