Nice analysis, I remember combining tf-idf with many other NLP techniques for my final year project and it was super easy to implement with nlp_compromise's tok

Nice analysis, I remember combining tf-idf with many other NLP techniques for my final year project and it was super easy to implement with nlp_compromise's tokenize(): nlp.tokenize(text, { dont_combine: true }).reduce(function(sentences, nlp, key){ var words = nlp.tokens.reduce(function(tokens, token){ // ... }); // ... });