Generating Responses Expressing Emotion in an Open-domain Dialogue System
- Chenyang Huang
- Osmar R. Zaiane, University of Alberta (Database)
Neural network-based Open-ended conversational agents automatically generate responses based on predictive models learned from a large number of pairs of utterances. The generated responses are typically acceptable as a sentence but are often dull, generic, and certainly devoid of any emotion. In this paper we present neural models that learn to express a given emotion in the generated response. We propose four models and evaluate them against 3 baselines. An encoder-decoder framework-based model with multiple attention layers provides the best overall performance in terms of expressing the required emotion. While it does not outperform other models on all emotions, it presents promising results in most cases.
Citation
C. Huang, O. Zaiane. "Generating Responses Expressing Emotion in an Open-domain Dialogue System". CONVERSATIONS: 2nd International Workshop on Chatbot Research, St. Petersburg, Russia, October 2018.| Keywords: | Open-domain dialogue generation, Emotion, Seq2seq, Attention mechanism |
| Category: | In Conference |
| Web Links: | Webdocs |
BibTeX
@incollection{Huang+Zaiane:18,
author = {Chenyang Huang and Osmar R. Zaiane},
title = {Generating Responses Expressing Emotion in an Open-domain Dialogue
System},
booktitle = {CONVERSATIONS: 2nd International Workshop on Chatbot Research},
year = 2018,
}Last Updated: September 10, 2020Submitted by Sabina P
