I decided to do some research in sentiment analysis and classification. A colleague of mine made some experiments and got me excited too.
I collected a dataset of nearly 5000 English poems by different (classic) poets and I will try to analyze them to see, whether I can get something useful out of them :)
By reading tons of scientific papers regarding sentiment analysis, I feel I could start writing a book, like Sentiment analysis for dummies :) Probably publishers of the Dummies book series would sue my pants off. Well, maybe I'll write the book anyway, but change the title.
Just kidding. I have to work on my PhD thesis.