arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1809.04276cs.CL
收藏

用于神经响应生成的检索增强对抗训练

Retrieval-Enhanced Adversarial Training for Neural Response Generation

Microsoft Research Asia(微软亚洲研究院);Harbin Institute of Technology(哈尔滨工业大学)

Qingfu Zhu, Lei Cui, Weinan Zhang, Furu Wei, Ting Liu

摘要:

对话系统通常建立在基于生成或基于检索的方法上,但它们不能从不同模型的优点中获益。在本文中,我们提出了一种用于神经响应生成的检索增强对抗训练(REAT)方法。与现有方法不同的是,REAT方法在对抗训练范式方面利用编码器-解码器框架,同时利用基于检索的系统中的N-最佳响应候选来构建模型。在大规模公共基准数据集上的实证研究表明,REAT方法的性能显着优于香草Seq 2Seq模型以及传统的对抗训练方法。

英文摘要:

Dialogue systems are usually built on either generation-based or retrieval-based approaches, yet they do not benefit from the advantages of different models. In this paper, we propose a Retrieval-Enhanced Adversarial Training (REAT) method for neural response generation. Distinct from existing approaches, the REAT method leverages an encoder-decoder framework in terms of an adversarial training paradigm, while taking advantage of N-best response candidates from a retrieval-based system to construct the discriminator. An empirical study on a large scale public available benchmark dataset shows that the REAT method significantly outperforms the vanilla Seq2Seq model as well as the conventional adversarial training approach.

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。

↑