Oct 2026
Spawning tasks from chitchatting - reinforcement learning-based policy
This research explores the transition from chitchatting in open-domain dialogue (ODD) to task-oriented dialogue (TOD) within a unified framework, and proposes a framework where chitchat is treated as a selectable action by the dialogue policy, alongside task-oriented actions.
Atheer Algherairy, Moataz Ahmed
· Neural computing & applicati... · 0 citations