'Emotion Analysis in Social Media: Corpus Annotation and Predictive Models' is the title of the next seminar in the series Linguistics Circle. The speaker is Dr Roman Klinger, University of Stuttgart.
The seminar will be held on Friday 9 March at 12:00 in Room 156 Gateway Building (GW156), University of Malta Msida Campus
Emotion analysis in text is an extension to sentiment analysis in the sense that classes to be predicted are not only positive and negative, but (for instance) joy, anger, fear, surprise, disgust, and sadness.
In the first part of this talk, Dr Roman Klinger will report on a corpus annotation study on Twitter and results of state-of-the-art models known from sentiment analysis. In the second part, he will highlight some remaining challenges, including multimodal information (images and text) and the meaning of modifiers (which emotion is "I am not angry"?).