Be sure to review our Idea Submission Guidelines for more information!
Submission GuidelinesHello,
Currently, the Sentiment analysis tool will score textual data by use of a Compound_Sentiment_Score, which is effectively a score from -1(negative) to 1(positive).
The tool is great, in that it has a built-in classification function so that we can say that anything under -0.5 we want to label as negative, and anything over 0.5 we want to label as positive.
However currently, the defaults are set to:
Which effectively creates a negative-weighted classification, as anything from -1 to -0.1 are classified as negative. -1 to 0.5 are labelled as neutral, and 0.5 to 1 are labelled as positive.
My suggestion would be to change the default max negative classification to 0.5, to level the weighting of the tools scoring.
Thanks,
TheOC
Pour ajouter un commentaire ici, vous devez être inscrit. Si vous êtes déjà inscrit, connectez-vous. Dans le cas contraire, inscrivez-vous puis connectez-vous.