How does the DoxCycle Classifier “learn”?
When you add a document to the classifier (see below), DoxCycle stores information about the document and relates it to the document type you assigned. The information may include words on the page, layout of the document, and more. The classifier uses this information to analyse new documents and automatically assign a type.
(Beware that sensitive information from the document may be stored during this process.)
When automatically assigning a type, the classifier weighs characteristics of an individual document against what it knows about all the document types. When the classifier calculates a “probable match” with one of those types it assigns the type.
By adding a document to the classifier, you are affecting the weight of the recognized characteristics and the calculation for matching documents. It modifies the probability that the document you added will be recognized, but it also affects the probability of recognizing other document types.