Black-Box Emotion Detection: On the Variability and Predictive Accuracy of Automated Emotion Detection Algorithms
Type
conference contribution
Date Issued
2020-10-02
Author(s)
Abstract (De)
The current research demonstrates considerable variability in predictive accuracy across major emotion detection systems (such as Google ML or Microsoft Cognitive Services) with lower (higher) classification accuracy for negative (positive) discrete emotions. We provide two modeling strategies to improve prediction accuracy by either combining feature sets or using ensemble methods.
Language
English
Keywords
emotion
recognition
ai
HSG Classification
contribution to scientific community
Event Title
ACR Conference 2020
Event Location
Paris
Event Date
1-4th October
Division(s)
Eprints ID
261517