The workflow starts reads ten famous fairy tales of the Brothers Grimm and two dictionaries containing the names or part of the names. Branch 1: The words of the first dictionary are searched by the Dictionary Tagger node in the documents. Found terms in the documents are tagged with PERSON tags. Document terms and dictionary words are compared case insensitive by exact match. Then the tagged terms are filtered and a bag of words is created. Branch 2: The words of the second dictionary are search in the documents by the Dictionary tagger. Found terms in the documents are tagged with PERSON tags. Document terms and dictionary words are compared case insensitive by non-exact match, meaning that a document term is tagged if it containes a word of the dictionary. Then the tagged terms are filtered and a bag of words is created.