A survey of some Italian literature works using sentiment analysis

Abstract

The article describes the use of software, written by the author, to analyze some Italian literature works from the last century by using sentiment analysis. The software is lexicon-based, with a sentiment Italian dictionary including about 30,000 positive and negative words. To set the sentiment of each word, a scale of points ranging from 0 to 100 was introduced. To check the overall sentiment in literary works, the software makes use of a new parameter, the index of positivity. It is found that sentiment analysis is an efficient way to detect the tendency of opinions (positive or negative) in literature works. The results of this analysis coincide with the critics on the classification of authors’ tendencies. The index of positivity could be useful also for bookstores: the customers, checking the index, would expect in advance, which is the overall tendency in a literary work. In this case, it would be easier to select a certain kind of work, according to the taste or the wish of customers.