Hello,
After my sentiment analysis, I realized that I have many text classified as positive but have negative sentiment score. Not sure how it is possible? I attached the screenshot of the file I exported from Knime.
Thanks in advance,
Begum
Hello,
After my sentiment analysis, I realized that I have many text classified as positive but have negative sentiment score. Not sure how it is possible? I attached the screenshot of the file I exported from Knime.
Thanks in advance,
Begum
Take a look at the Rule Engine node in your workflow. I suspect it looks something like this:
Note that the POS / NEG classification is being assigned relative to the average sentiment score, NOT zero.
Thanks for this, yes I checked it, and it looks exactly like this. However, there is one point I got still confused. So, sentiment score is calculated for each document as: (number of positive words - number of negative words) / total number of words. So, when we have more positive words in the text, the sentiment classification should be positive, right? But when I used this scoring method, there are some sentences, even though the number of negative words is more than the number of positive words, the sentiment classification is still positive. How it is possible?
Good catch - I guess this is because the scoring method in the Math Formula node and the Rule Engine classification don’t match up right. So you’d just need to modify one or the other for the approach you prefer.
This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.