It's easy to spot the trend in Chapter 4, Gensim - Vectorizing Text and Transformations and n-grams, Chapter 5, POS-Tagging and Its Applications, and Chapter 6, NER-Tagging and Its Applications - all of which choose spaCy as the preferred implementation, not just for the accuracy and speed, but for the way it naturally fits into our text analysis pipelines. We still discussed the other Python libraries available to perform the task, and we will do the same for dependency Parsing.
As usual, we will start with NLTK, which provides the most options regarding parsing methods, but unlike the previous cases, a not so intuitive API and one where we are forced to pass our own grammar for effective results. It is not our purpose to learn grammars before we run computational linguistic algorithms, and this is another reason we will always prefer spaCy for...
