arXiv 1702.067032017-02-22cs.CL

用于控制对话生成中的特异性的数据蒸馏

Data Distillation for Controlling Specificity in Dialogue Generation

Stanford University（斯坦福大学）

Jiwei Li, Will Monroe, Dan Jurafsky

摘要：

人们在不同的情况下说话的具体程度不同。取决于他们的知识、对话者、情绪等。会话代理应该具有这种能力，并且知道什么时候应该具体，什么时候应该笼统。我们提出了一种方法，使一个神经网络-基于会话代理这种能力。我们的方法涉及在\emph{data distillation}和模型训练之间交替：删除与上一轮训练的模型最常产生的响应最接近的训练样本，然后在剩余数据集上重新训练模型。用不同程度的数据提炼训练的对话生成模型表现出不同程度的特异性。然后，我们训练一个强化学习系统，用于在这个生成模型池中进行选择，为给定的输入选择最佳的特异性水平。与未经蒸馏训练的原始生成模型相比，除了根据上下文适当调整特异性外，所提出的系统还能够生成更有趣和更高质量的响应。我们的研究构成了更广泛方法的一个具体案例，该方法涉及从单个数据集训练多个子系统，这些数据集通过希望建模的特定属性的差异来区分。我们表明，从这样一组子系统，可以使用强化学习来构建一个系统，在测试时将其输出量身定制为不同的输入上下文。

英文摘要：

People speak at different levels of specificity in different situations. Depending on their knowledge, interlocutors, mood, etc.} A conversational agent should have this ability and know when to be specific and when to be general. We propose an approach that gives a neural network--based conversational agent this ability. Our approach involves alternating between \emph{data distillation} and model training : removing training examples that are closest to the responses most commonly produced by the model trained from the last round and then retrain the model on the remaining dataset. Dialogue generation models trained with different degrees of data distillation manifest different levels of specificity. We then train a reinforcement learning system for selecting among this pool of generation models, to choose the best level of specificity for a given input. Compared to the original generative model trained without distillation, the proposed system is capable of generating more interesting and higher-quality responses, in addition to appropriately adjusting specificity depending on the context. Our research constitutes a specific case of a broader approach involving training multiple subsystems from a single dataset distinguished by differences in a specific property one wishes to model. We show that from such a set of subsystems, one can use reinforcement learning to build a system that tailors its output to different input contexts at test time.

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。

摘要：

英文摘要：

相关论文