Books were downloaded from ProjectGutenberg.com through a custom-made webscarping/querying routine written in Python.
Sentiment Analysis was carried out using:
a) Naive Bayes Classifier trained using the twitter_samples dataset,
b) VADER Polarity Scores
c) NRC Lexicon