Empathic dialogue system based on emotions extracted from tweets
summary
System that pulls news headlines and related Tweets then classifies Tweets and utterances into positive negative neutral and picks a Tweet whose emotion matches the user to form a response. Claims that matching tweet emotion to user emotion raises perceived empathy for explicitly positive or negative user turns. Evaluation builds a working chat prototype and runs a pilot with 54 participants and a 64‑participant main study; results show the proposed method wins when user emotion is positive or negative but not for neutral turns and random replies score highest overall.
AI-written summary · about
chapters (11)
- 0:00Introduction to empathic dialogue
- 1:09Limitations of existing systems
- 2:19Base system overview
- 3:05Proposed empathic system
- 4:13Preliminary experiment results
- 6:06Main experiment setup
- 7:11Empathy measurement results
- 8:40Analysis of positive and negative empathy
- 9:26Neutral empathy challenges
- 10:33Conclusion and future work
- 11:21Q&A session
related talks
39mThe art and science of making computers understand sentiment and emotion
01h 23mBeyond Sentiment Emotion Mining with Python and machine learning
46mAI That Reads Emotions: What It Means for Data Privacy, Consent, Your Teams
11mAmbient Intelligence and Emotionally Intelligent Devices
15mRecognizing Emotional Expression in Game Streams
01h 05mSocial Text, Sentiment, and Tone Analysis
32mDecoding Emotions from Art to Memes
27mSentiment analysis of tweets in Polish language using deep...