Enhancing Target-Guided Proactive Dialogue Systems via Conversational Scenario Modeling and Intent-Keyword Bridging

2026-05-12

AI Summary

This work addresses the challenge that existing goal-oriented proactive dialogue systems struggle to dynamically model conversational contexts and intent-related keywords, often generating responses misaligned with authentic dialogue dynamics. To overcome this limitation, the authors propose a novel approach that jointly leverages user profiles and domain knowledge to construct a dynamic dialogue scene representation. The method introduces a dynamic scene bias and an intent-keyword bridging mechanism to proactively predict intent keywords for upcoming turns, thereby enabling fine-grained, high-level guidance of system utterance generation. Experimental results demonstrate that the proposed framework significantly enhances the system’s proactiveness, fluency, and informativeness in both automatic and human evaluations, effectively narrowing the gap between system-generated and human-like dialogues.

Abstract

A target-guided proactive dialogue system aims to steer conversations proactively toward pre-defined targets, such as designated keywords or specific topics. During guided conversations, dynamically modeling conversational scenarios and intent keywords to guide system utterance generation is beneficial; however, existing work largely overlooks this aspect, resulting in a mismatch with the dynamics of real-world conversations. In this paper, we jointly model user profiles and domain knowledge as conversational scenarios to introduce a scenario bias that dynamically influences system utterances, and employ intent-keyword bridging to predict intent keywords for upcoming dialogue turns, providing higher level and more flexible guidance. Extensive automatic and human evaluations demonstrate the effectiveness of conversational scenario modeling and intent keyword bridging, yielding substantial improvements in proactivity, fluency, and informativeness for target-guided proactive dialogue systems, thereby narrowing the gap with real world interactions.

Problem

• target-guided dialogue
• proactive dialogue system
• conversational scenario
• intent keyword
• dialogue dynamics

Innovation

• conversational scenario modeling
• intent-keyword bridging
• target-guided dialogue
• proactive dialogue system
• dynamic utterance generation