Get Started
Home
Topics
Search
Library
Research questionHow can controlled synthetic dialogues satisfy intended emotion and intent while remaining natural to human readers?Synthetic service dialogues must express specified intent and emotion while maintaining coherent, fluent exchanges. Automatic acceptance can improve without ensuring that people perceive the emotional realization as appropriate or the conversation as natural.
AI
Evaluation & Benchmarks
Machine Learning
Natural Language Processing
Latest papersRecent research connected to this question, newest first.PragAlign: Feedback-Guided Pragmatic Alignment for Controlled Synthetic Dialogue GenerationThe source studies dialogues conditioned on service context, target intent, target emotion, and auxiliary trait-style controls. Its generate–evaluate–revise framework uses an LLM-based evaluator for intent, emotion, coherence, fluency, and overall quality, with evidence from 800 matched specifications and human judgments of 1,200 generated dialogues. Human recognition of intent and dialogue flow was stronger than the stability of emotion appropriateness; broader independent human-perceived quality remains unresolved.research paper · Sep 2, 2026
Related questions
How can empathetic dialogue systems control response stance toward a speaker’s affect?How can speech synthesis produce natural dubbing and full-duplex dialogue without forced alignment or explicit duration prediction?How can full-duplex dialogue models learn natural acoustic turn-taking without degrading semantic responses?How should conversational foundation models be evaluated when latency and generation speed shape user experience?