As you can see from the above responses, there are two common data points available in both the responses :
Sentiment Polarity:
It is the score generated by Symbl.ai’s algorithm to determine the polarity of this score . The overall polarity can range from -1 to +1 .
Sentiment Suggested :
This is the suggested sentiment for the sentence by the Symbl’s sentiment models . For the above example it is “Positive”.
Polarity Ranges and corresponding sentiment suggestion:
As Symbl’s sentiment polarity ranges from -1 to +1 , given below are the suggestions for the sentiment .
Polarity | Suggested Sentiment |
---|---|
-1.0 => x > -0.3 | negative |
-0.3 => x <= 0.3 | neutral |
0.3 < x <= 1.0 | positive |
The above ranges are both suggestions as well as indications, for example, a message with -0.4 score is negative , but one with -0.9 is highly negative .
This way symbl has given power to the developers and product personas to further classify the sentiments as per their use case .