Proactive Dialogue Model with Intent Prediction
For dialogue system developers, this provides a simple, model-agnostic way to improve efficiency in multi-intent conversations, though the gains are incremental.
The paper introduces a lightweight intent-transition prior injected into the system prompt to make dialogue models proactive, reducing redundant interactions in multi-intent settings. The method improves Coverage AUC from 0.742 to 0.856 and reduces turns to 75% intent coverage from 3.95 to 2.73.
Dialogue models are inherently reactive, responding to the current user turn without anticipating upcoming intents, which leads to redundant interactions in multi-intent settings. We address this limitation by introducing a lightweight intent-transition prior derived from dialogue data and injected into the system prompt at inference time. We instantiate this prior using a Temporal Bayesian Network (T-BN) trained on per-turn intent annotations in MultiWOZ 2.2. The T-BN achieves Recall@5 = 0.787 and MRR = 0.576 on 1,071 held-out USER-turn pairs. In a ground-truth replay over 200 dialogues, BN-guided generation improves Coverage AUC from 0.742 to 0.856 and reduces the number of turns required to reach 75% intent coverage from 3.95 to 2.73. These results show that lightweight intent-transition guidance enables more proactive and efficient dialogue behavior without modifying the underlying language model.