Is the system using "1-grams"? i.e., is it not context aware? For instance, it's picking up "can’t blame" as anger, taking only into account "blame".
It's also picking up the "inferior" in "Our product is in no way inferior" as being negative - but given context, the word is clearly not being used in a negative way.
Is the system using "1-grams"? i.e., is it not context aware? For instance, it's picking up "can’t blame" as anger, taking only into account "blame".
It's also picking up the "inferior" in "Our product is in no way inferior" as being negative - but given context, the word is clearly not being used in a negative way.