HIGHLIGHTS
- who: Deeksha Varshney and colleagues from the BANGLADESH , Samsung Research India have published the research: EmoKbGAN: Emotion controlled response generation using Generative Adversarial Network for knowledge grounded conversation, in the Journal: PLOS ONE of 20/Oct/2021
- what: Since MLE loss evaluates the sentences at the word level the authors focus on sentence-level judgment for the training purposes. The primary motivation behind the current work is to generate more socially engaging conversations, by considering emotion labels and relevant-facts associated with multi-turn conversations. The authors evaluate the proposed model on two benchmark datasets . . .
If you want to have access to all the content you need to log in!
Thanks :)
If you don't have an account, you can create one here.