Advancing Open-Domain Conversational Agents - Designing an Engaging System for Natural Multi-Turn Dialogue

Islam A. Hassan, Yvette Graham


Abstract
This system paper describes our conversational AI agent developed for the SCI-CHAT competition. The goal is to build automated dialogue agents that can have natural, coherent conversations with humans over multiple turns. Our model is based on fine-tuning the Snorkel-Mistral-PairRM-DPO language model on podcast conversation transcripts. This allows the model to leverage Snorkel-Mistral-PairRMDPO’s linguistic knowledge while adapting it for multi-turn dialogue modeling using LoRA. During evaluation, human judges will converse with the agent on specified topics and provide ratings on response quality. Our system aims to demonstrate how large pretrained language models, when properly adapted and evaluated, can effectively converse on open-ended topics spanning multiple turns.
Anthology ID:
2024.scichat-1.8
Volume:
Proceedings of the 1st Workshop on Simulating Conversational Intelligence in Chat (SCI-CHAT 2024)
Month:
March
Year:
2024
Address:
St. Julians, Malta
Editors:
Yvette Graham, Qun Liu, Gerasimos Lampouras, Ignacio Iacobacci, Sinead Madden, Haider Khalid, Rameez Qureshi
Venues:
SCI-CHAT | WS
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
75–79
Language:
URL:
https://aclanthology.org/2024.scichat-1.8
DOI:
Bibkey:
Cite (ACL):
Islam A. Hassan and Yvette Graham. 2024. Advancing Open-Domain Conversational Agents - Designing an Engaging System for Natural Multi-Turn Dialogue. In Proceedings of the 1st Workshop on Simulating Conversational Intelligence in Chat (SCI-CHAT 2024), pages 75–79, St. Julians, Malta. Association for Computational Linguistics.
Cite (Informal):
Advancing Open-Domain Conversational Agents - Designing an Engaging System for Natural Multi-Turn Dialogue (Hassan & Graham, SCI-CHAT-WS 2024)
Copy Citation:
PDF:
https://aclanthology.org/2024.scichat-1.8.pdf