Sentiment analysis categorizes a piece of text in two categories: "appreciation" and "concern". It's possible that text can belong to only one category, both categories, or neither of them.Some examples:
text "It's Tuesday" would be categorized as "not appreciation" and "not concern".
text "The food was bad but I really enjoyed the movie." would get categorized into both "appreciation" and "concern".
Admin users can provide training data by using the sentiment analysis in full screen and then choosing from example texts what they think is the correct sentiment. This happens across whole Howspace - all workspaces by all customers. This training data is used to improve the sentiment model. Training data is currently being curated by Howspace team.