Boosting Dialog Response Generation
Boosting Dialog Response Generation
复制标题
DOI：
10.18653/v1/p19-1005
复制
发表时间：
2019-07
期刊：
影响因子：
--
通讯作者：
Wenchao Du;A. Black
中科院分区：
文献类型：
--
作者：
Wenchao Du;A. Black
神经模型已经成为对话响应生成的最重要的方法之一。然而，他们仍然倾向于在语料库中生成最常见和最通用的响应。为了解决这个问题，我们设计了一个迭代的训练过程和集成方法的基础上提升。我们将我们的方法与不同的训练和解码范式相结合作为基础模型，包括基于互信息的解码和奖励增强的最大似然学习。实证结果表明，我们的方法可以显着提高的多样性和相关性的所有基础模型生成的响应，客观的测量和人类的评价支持。
Neural models have become one of the most important approaches to dialog response generation. However, they still tend to generate the most common and generic responses in the corpus all the time. To address this problem, we designed an iterative training process and ensemble method based on boosting. We combined our method with different training and decoding paradigms as the base model, including mutual-information-based decoding and reward-augmented maximum likelihood learning. Empirical results show that our approach can significantly improve the diversity and relevance of the responses generated by all base models, backed by objective measurements and human evaluation.