Collaborate

Could we implement extra filtering options for sentiment to look for mentions coded as unknown/processing, similar to how we do with location data? This would be helpful for clients to get a sense of what's missing from sentiment in large data sets