thanks, the sentiment in these graphs seem more positive in comparison. Did you run the sentiment on the whole corpus? What did that look like?