Sentiment Inaccuracy Suspected
(Doc ID 2569160.1)
Last updated on OCTOBER 15, 2020
Applies to:Oracle Social Engagement and Monitoring Cloud Service - Version N/A to N/A [Release V1]
Information in this document applies to any platform.
While monitoring one topic around employee experience, user was expecting the sentiment to be more accurate as the reviews contain clearly identifiable negative or positive keywords.
They are aware the automatically generated sentiment will have a certain percentage of (in)accuracy, however when looking at some of the “neutral” mentions, they think that the keywords in it are clearly either positive or negative, and there is also not a combination of both positive or negative (which could lead to confusion). Great, fear, rewarding, unhealthy, inconsistent, productive, poor, dissatisfied, positive are a few of those keywords.
The topics are set to Whole Post Mention.
To view full details, sign in with your My Oracle Support account.
Don't have a My Oracle Support account? Click to get started!
In this Document