The documentation overall is very light and some of it is inaccurate, for instance new SentimentAnalyzer({ language: 'en' }).getSentiment() does not return a Promise even though the documentation shows it used with .then() in all examples.

There is also no examples of what the score and comparitive values represent or a small overview of how some of the processes work and what the results mean. Again with Sentiment Analysis I can guess that scores > 0 are positive and < 0 are negative, but what are the value ranges this sits between? If I get back a value of 8 is that a lot more positive than 0.25, or only marginally positive/almost neutral?

What are the comparative values for, what do they represent. A little time spent on this would make it a lot easier to understand and work with as it seems like a great way to cleanly and easily introduce NLP to a development team.

Kind of also ties in with #121