CausalDialogue: Modeling Utterance-level Causality in Conversations
This work addresses the challenge of improving dialogue generation for AI systems, but it is incremental as it builds on existing methods with a new causality-focused approach.
The paper tackled the problem of neural conversation models lacking natural chat capabilities by modeling utterance-level causality, where user utterances are causes and responses are effects, and proposed the ExMATE method to enhance causality in training, resulting in improved diversity and agility in responses on the new CausalDialogue dataset.
Despite their widespread adoption, neural conversation models have yet to exhibit natural chat capabilities with humans. In this research, we examine user utterances as causes and generated responses as effects, recognizing that changes in a cause should produce a different effect. To further explore this concept, we have compiled and expanded upon a new dataset called CausalDialogue through crowd-sourcing. This dataset includes multiple cause-effect pairs within a directed acyclic graph (DAG) structure. Our analysis reveals that traditional loss functions struggle to effectively incorporate the DAG structure, leading us to propose a causality-enhanced method called Exponential Maximum Average Treatment Effect (ExMATE) to enhance the impact of causality at the utterance level in training neural conversation models. To evaluate the needs of considering causality in dialogue generation, we built a comprehensive benchmark on CausalDialogue dataset using different models, inference, and training methods. Through experiments, we find that a causality-inspired loss like ExMATE can improve the diversity and agility of conventional loss function and there is still room for improvement to reach human-level quality on this new dataset.