I am beginner in natural language processing and my goal is to find a way to score sentences based on their emotional tension. More specifically, I would like to know to what degree a sentence transmits wish, hate, or fear. I looked at some studies on sentiment analysis, but haven't seen any relevant result. Most of the them are on negativity or positivity of a sentence.


1 Answer 1


Consider multi-label classification, instead of a binary sentiment your dataset would have a level/degree/probability target for each emotional label of whatever dimension size you want.

In the context of neural networks, it's just a matter of replacing softmax with sigmoid loss layer.

If getting such large dataset with desired supervised labels is hard or not possible, one idea may be to use models trained on any sentiment datasets out there (there are even some multi-dimensional too) and apply additional un(semi)-supervised learning techniques to extract what you want.

  • $\begingroup$ Yeah theoritically it's easy to explain. How are you gonna get the data? or rather, how are you going to assign emotional states to speech with a certain degree of confidence, it's hard to label speech with the emotional state of a person since there are many nuances to a speech. $\endgroup$
    – Blenz
    Commented Jun 25, 2019 at 14:08
  • $\begingroup$ @Blenz Getting supervised data for this task is a totally different question, maybe I've just mistakenly assumed OP is asking for a beginner modeling advice. $\endgroup$
    – swish
    Commented Jun 25, 2019 at 15:43
  • $\begingroup$ yes, what he's asking hasn't been done until now i think, haven't seen papers on this recently but could be something soon to be out there. But in theory yes, the structure and the algorithms for it exist, the problem is data $\endgroup$
    – Blenz
    Commented Jun 25, 2019 at 15:54
  • $\begingroup$ @Blenz: Some emotion-score labelled data: saifmohammad.com/WebPages/EmotionIntensity-SharedTask.html $\endgroup$
    – Ben Reiniger
    Commented Jun 25, 2019 at 16:33
  • $\begingroup$ @BenReiniger thanks for the link $\endgroup$
    – Blenz
    Commented Jun 25, 2019 at 16:42

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.