BriefGPT.xyz
Oct, 2023
面向任务导向对话系统的语境数据增强
Contextual Data Augmentation for Task-Oriented Dialog Systems
HTML
PDF
Dustin Axman, Avik Ray, Shubham Garg, Jing Huang
TL;DR
我们开发了一种新型的对话扩充模型,通过完整的对话上下文生成用户的回合,并通过语言模型的新提示设计和输出重新排序,所生成对话可直接用于训练下游对话系统,在常见的基准数据集MultiWoZ和SGD上,展示了我们的对话扩充模型生成高质量对话并使对话成功率较基准线提高多达8%。
Abstract
Collection of annotated dialogs for training
task-oriented dialog systems
have been one of the key bottlenecks in improving current models. While
dialog response generation
has been widely studied on the agent si
→