Discourse AI感情分析の新機能:管理者向け新レポート登場

Is there a way to modify an analysis that is incorrect?

For example, we have a section of our community for guides, which use a “issue/problem addressed & solution” format. It’s flagging all these as negative, presumably because they’re describing an issue—but I think they are far from negative when it comes to sentiment.

「いいね!」 2