Will I Sound Like Me? Improving Persona Consistency in Dialogues through Pragmatic Self-Consciousness
This work addresses the issue of maintaining persona consistency in dialogue systems, which is crucial for creating more reliable and human-like conversational agents, though it is incremental as it builds on existing frameworks.
The paper tackles the problem of persona inconsistency in dialogue agents by introducing a pragmatic self-consciousness mechanism based on the Rational Speech Acts framework, which reduces contradictions and improves consistency without requiring additional training labels or modules.
We explore the task of improving persona consistency of dialogue agents. Recent models tackling consistency often train with additional Natural Language Inference (NLI) labels or attach trained extra modules to the generative agent for maintaining consistency. However, such additional labels and training can be demanding. Also, we find even the best-performing persona-based agents are insensitive to contradictory words. Inspired by social cognition and pragmatics, we endow existing dialogue agents with public self-consciousness on the fly through an imaginary listener. Our approach, based on the Rational Speech Acts framework (Frank and Goodman, 2012), can enforce dialogue agents to refrain from uttering contradiction. We further extend the framework by learning the distractor selection, which has been usually done manually or randomly. Results on Dialogue NLI (Welleck et al., 2019) and PersonaChat (Zhang et al., 2018) dataset show that our approach reduces contradiction and improves consistency of existing dialogue models. Moreover, we show that it can be generalized to improve context-consistency beyond persona in dialogues.