A question regarding Sentiment score

Hello there:



  I have seen that  in some courses  in Quantra, there is a library on python called Vader, that is used for  scoring news…



Now I was wondering… if this scoring could be used as an indicator when considering to enter a trade… that means… if this scoring is near 1  , let's say above 0,8 ;  It can be considered as an argument to go long on a particular stock  (of course this would not be the only one…)



Also I wonder  How does Vader actually come with this number…??  what about all the scores?,  that is  the neg, the neutral and the positive ??





Thanks

Hi Ghery,



Yes, you can use it as an additional signal to enter trades. Treating values above 0.8 as a bullish (long) indicator and scores below -0.8 as bearish makes sense. This compound score is a normalised metric that combines Vader’s individual positive, negative, and neutral sentiment scores, weighing them to give an overall sentiment ranging from -1 to 1.



Vader's scores are calculated based on a lexicon and rule-based approach:

  1. Positive, Negative, Neutral Scores: Each score represents the proportion of words in the text that are classified as positive, negative, or neutral. These are calculated based on a dictionary of words that Vader knows to generally convey certain sentiments.
  2. Compound Score: This is the primary score to use, as it integrates the other three. It is a weighted sum of these scores, normalised between -1 and 1, representing the text’s overall sentiment intensity.



    If you want to learn more about Vader you can refer to this paper.



    Hope this helps!



    Thanks

    Rushda