Not bad, but getting the data is one thing, but this screams for a "so what?" question. Interpreting data and making sense of it is where the real analysis takes place. For example, you would need to compare what makes Tolkien books stand out from the rest of the literature and so on in order to drive observations and tentative conclusions.
This being said, great work and hopefully this is just the beginning of a continued effort to better understand Tolkien's work.
I disagree; I find the 'pure data' of this much more compelling than some arbitrary speculation about the meaning of the results.
For example, you could speculate that the sentiment analysis shows points of tension and peace in the story, and investigate that hypothesis... but is writing a blog about that, what situations it works in, and what situations it doesn't work in and coming up with a 'well, maybe...' kind of result what interesting? Not to me.
...but providing me with the tools to investigate that myself?
This is not really "pure" data as there could be tons of other analysis done on the text that one can do and you do not get the raw results, and the tools to get the data and do further analysis are not provided (AFAIK). Why did the author just limit itself to those, or why did those matter more than the other ? What is the purpose of getting the data in the first place? Usually getting data is a mean to answer a question, you do not generate data just for the purpose of the data itself. Hence my "so what" question.
I was looking at the sentiment analysis tab of the project. I really wonder how the algorithm works. It seems that for almost all pages and most books the sentiment is either negative or neutral. The absence of positive sentiment can be observed immediately.
This being said, great work and hopefully this is just the beginning of a continued effort to better understand Tolkien's work.