We should fine tune a sentiment model so that it recognizes things that recognizes hidden bias. Like
Russia Likely Lost 100 Combat Planes In Ukraine War, Says U.K.
is in fact positive but the naive sentiment checkers think this is sad.
I wonder if we should make a kind of dictionary of topics again based on sentiment? That would be weird but. I don't know. Could come up with some interesting dimensional analysis.
We should fine tune a sentiment model so that it recognizes things that recognizes hidden bias. Like
is in fact positive but the naive sentiment checkers think this is sad.
I wonder if we should make a kind of dictionary of topics again based on sentiment? That would be weird but. I don't know. Could come up with some interesting dimensional analysis.