sentiment prediction formula


I am using the workflow for Lexicon Based Sentiment Analysis and I am trying to understand the formula that was used in this workflow for sentiment prediction. I attached the screenshot of it and would be glad if someone can help me to understand this better. In addition to that, what I want to do is, I want to assign “POSITIVE” sentiment for all sentiment scores higher than “0”, assign “NEGATIVE” sentiment for all scores lower than “0” and all else assign “NEUTRAL”. How can I can do that? Thanks in advance for your help.

The way you can read this is:

If the Sentiment Score on this row is greater than or equal to the overall average of the Sentiment Score, then assign this row a label of “POS”. Otherwise, set it to “NEG”.

To make your desired change, just replace $${DMean(Sentiment Score)}$$ with 0.

Thank you very much. Is there an academic citation I can use for this calculation? I will use it in my working paper and I am looking for a citation that use this logic to predict the sentiments.

For this particular calculation, probably not. If you wish to cite KNIME more generally, check here:

1 Like

That would work too! Thanks!

1 Like