论文标题
改善响应与人格事实之间的事实一致性
Improving Factual Consistency Between a Response and Persona Facts
论文作者
论文摘要
响应产生的神经模型产生的响应在语义上是合理的，但不一定与描述说话者角色的事实一致。这些模型经过全面监督的学习培训，目标功能几乎无法捕获事实的一致性。我们建议通过强化学习和有效的奖励功能来微调这些模型，从而明确捕捉响应和人物事实之间的一致性以及语义上的合理性。我们对人为ACHATAT语料库的自动和人类评估证实，我们的方法提高了响应速度，这些响应速率与人为事实相一致，同时保留了响应的语言质量。
Neural models for response generation produce responses that are semantically plausible but not necessarily factually consistent with facts describing the speaker's persona. These models are trained with fully supervised learning where the objective function barely captures factual consistency. We propose to fine-tune these models by reinforcement learning and an efficient reward function that explicitly captures the consistency between a response and persona facts as well as semantic plausibility. Our automatic and human evaluations on the PersonaChat corpus confirm that our approach increases the rate of responses that are factually consistent with persona facts over its supervised counterpart while retaining the language quality of responses.
