Predicting various types of sentiment for textual data sources is an intensely studied problem in linguistics and computing. Despite progress with computational solutions beyond deterministic lookup dictionaries, practitioners ...