1

I have been trying to search for a scale or classification metric to assign some emotional degree to VADER sentiment analysis beyond just positive, negative or neutral. I would really appreciate if someone can share their view or a resource to help classify VADER compound score along the following lines:

-0.75 to -1    -  Angry
-0.5 to  -0.74 -  Frustrated
-0.25 to -0.49 -  Unhappy
-0.01 to -0.24 -  Dissatisfied/Confused
0 to 0.25      -  Indifferent
0.26 to 0.5    -  Satisfied
0.51 to 0.75   -  Happy
0.76 to 1.0    -  Thrilled
Salik
  • 508
  • 6
  • 17
  • 35

1 Answers1

1

Vader is a library for sentiment analysis that uses lexicon. Classifying sentiments into compound scores is a different subject than classifying them to categories. I'm not sure how accurate would be for example to classify a negative sentiment with a score range of -0.75 to -1 to angry.

The sentiment might just be about something very sad so classifying it as "angry" would be pretty inaccurate. Regardless Vader does not support the feature you ask for and what you ask for might be connected to sentiments but it is a different field in NLP.

Nareg
  • 11
  • 3