Content for my website
For my final project at Ubiqum Code Academy I analysed a number of classic literature books with the aim of visualizing the content, or rather the sentiments (or the mood/ feelings if you will) associated with the text. The books I used for this analysis:
I chose two books with a rather upbeat vibe and two books on the rather sombre end of the spectrum (as you probably can already tell by the titles). Main reason being that I wanted to see how the various sentiment packages score the content of each book and inhowfar the outputs differ.
You’ll find a number of visualisations based on the output data under the links further below.
Sentiment analyses is one of the standard methods for analysing any kinds of texts with R. Typical applications for text analysis are for example analysing tweets, evaluating user reviews (e.g. on Amazon), gathering what newspapers are saying on specific topics or also for getting a picture of the trending topics within the webosphere as a whole (forums such as Reddit, blogs, instagram etc).
I’ll explain here more about the details of the project, the dictionaries I used etc. But bascially I wanted to see how far one can use the standard dictionaries within R (NRC, Bing, AFINN & SentimentR) to analyse content (in this case a number of well known books) and in how far the results differ.
Producing visualisations with Tableau made up a considerable part of the project, so without further ado, here they are:
For more details about the actual text analysis