syn_extractor using wordnet from nltk library, the script extract all tokens and then return a dataframe with the word and its relative synonymous, if the wordnet synset of a word is not empty