Knowledge Commons of Institute of Automation,CAS
Domain-Lifelong Learning for Dialogue State Tracking via Knowledge Preservation Networks | |
Liu QB(刘庆斌)1,2![]() ![]() ![]() ![]() ![]() ![]() ![]() | |
2021-11 | |
会议名称 | Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing |
会议日期 | 2021-11 |
会议地点 | online |
摘要 | Dialogue state tracking (DST), which estimates user goals given a dialogue context, is an essential component of task-oriented dialogue systems. Conventional DST models areusually trained offline, which requires a fixed dataset prepared in advance. This paradigmis often impractical in real-world applications since online dialogue systems usually involve continually emerging new data and domains. Therefore, this paper explores Domain-Lifelong Learning for Dialogue State Tracking(DLL-DST), which aims to continually train a DST model on new data to learn incessantly emerging new domains while avoiding catastrophically forgetting old learned domains.To this end, we propose a novel domain-lifelong learning method, called Knowledge Preservation Networks (KPN), which consists of multi-prototype enhanced retrospection and multi-strategy knowledge distillation, to solve the problems of expression diversity and combinatorial explosion in the DLL-DST task. Experimental results show that KPN effectively alleviates catastrophic forgetting and outperforms previous state-of-the-art lifelong learning methods by 4.25% and 8.27% of whole joint goal accuracy on the MultiWOZ benchmark and the SGD benchmark, respectively. |
DOI | 10.18653/v1/2021.emnlp-main.176 |
收录类别 | EI |
七大方向——子方向分类 | 自然语言处理 |
引用统计 | |
文献类型 | 会议论文 |
条目标识符 | http://ir.ia.ac.cn/handle/173211/46634 |
专题 | 多模态人工智能系统全国重点实验室_自然语言处理 |
通讯作者 | Zhao J(赵军) |
作者单位 | 1.National Laboratory of Pattern Recognition, Institute of Automation,Chinese Academy of Sciences, Beijing, China 2.School of Artificial Intelligence, University of Chinese Academy of Sciences, Beijing, China 3.Meituan, Beijing, China |
第一作者单位 | 模式识别国家重点实验室 |
通讯作者单位 | 模式识别国家重点实验室 |
推荐引用方式 GB/T 7714 | Liu QB,Cao PF,Liu C,et al. Domain-Lifelong Learning for Dialogue State Tracking via Knowledge Preservation Networks[C],2021. |
条目包含的文件 | 下载所有文件 | |||||
文件名称/大小 | 文献类型 | 版本类型 | 开放类型 | 使用许可 | ||
EMNLP_2021__Domain_L(711KB) | 会议论文 | 开放获取 | CC BY-NC-SA | 浏览 下载 |
除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。
修改评论