Corpus summary

The file texts.zip contains the ASCII files of the ebooks used in our study that were downloaded from the Gutenberg Project (https://www.gutenberg.org/) and Project Gutenberg Australia (http://gutenberg.net.au/index.html).The file vader_comp.zip contains the VADER compound sentiment value for each sentence on each book. These sequences are the primary inputs for our analysis.The file discrete_sentiment.zip contains the sequence of discrete sentiment values, when a threshold value is set to classified the sentiment. Convention: The positive sentences are marked as 0, the neutrals as 1, and the negatives by number 2.