I am going to classify tweets based on their positive and negative sentiments. A challenge is that there are some tweets extremely mean and bitter with highly positive words. Such tweets cannot be classified using ordinary models and classifiers.
I'm also studying emotion and understand the challenge you're talking about. I'm just now starting to dive into the corpus linguistics of my research, but from the readings I've done thus far, I plan to use the APPRAISAL method. You can see the website http://www.grammatics.com/appraisal/ for a preview. Please keep us posted about your work. It sounds interesting!
This chapter may interest you (that is, if you haven't read it yet):
Louw, B. (1993). Irony in the text or insincerity in the writer? The diagnostic potential of semantic prosodies. Text and technology: In honour of John Sinclair.
Jennifer Furlong Thank you very much for your answer. I will go through the web page. Also, Thanks for being interested in my work. Having advanced in my work, I share the results.