Knowledge Commons of Institute of Automation,CAS
A Unified Shared-Private Network with Denoising for Dialogue State Tracking | |
其他题名 | A Unified Shared-Private Network with Denoising for Dialogue State Tracking |
Liu QB(刘庆斌)1,2; He SZ(何世柱)1,2; Liu K(刘康)1,2; Liu SP(刘升平)3; Zhao J(赵军)1,2 | |
发表期刊 | JOURNAL OF COMPUTER SCIENCE AND TECHNOLOGY |
ISSN | 1000-9000(Print) /1860-4749(Online) |
2021-11 | |
卷号 | 36期号:6页码:1407-1419 |
产权排序 | 1 |
文章类型 | 期刊论文 |
摘要 | Dialogue state tracking (DST) leverages dialogue information to predict dialogues states which are generally represented as slot-value pairs. However, previous work usually has limitations to efficiently predict values due to the lack of a powerful strategy for generating values from both the dialogue history and the predefined values. By predicting values from the predefined value set, previous discriminative DST methods are difficult to handle unknown values. Previous generative DST methods determine values based on mentions in the dialogue history, which makes it difficult for them to handle uncovered and non-pointable mentions. Besides, existing generative DST methods usually ignore the unlabeled instances and suffer from the label noise problem, which limits the generation of mentions and eventually hurts performance. In this paper, we propose a unified shared-private network (USPN) to generate values from both the dialogue history and the predefined values through a unified strategy. Specifically, USPN uses an encoder to construct a complete generative space for each slot and to discern shared information between slots through a shared-private architecture. Then, our model predicts values from the generative space through a shared-private decoder. We further utilize reinforcement learning to alleviate the label noise problem by learning indirect supervision from semantic relations between conversational words and predefined slot-value pairs. Experimental results on three public datasets show the effectiveness of USPN by outperforming state-of-the-art baselines in both supervised and unsupervised DST tasks. |
关键词 | dialogue state tracking unified strategy shared-private network reinforcement learning |
DOI | 10.1007/s11390-020-0338-0 |
收录类别 | SCI |
语种 | 英语 |
WOS记录号 | WOS:000730175900011 |
七大方向——子方向分类 | 自然语言处理 |
引用统计 | |
文献类型 | 期刊论文 |
条目标识符 | http://ir.ia.ac.cn/handle/173211/46635 |
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 | He SZ(何世柱) |
作者单位 | 1.National Laboratory of Pattern Recognition, Institute of Automation, Chinese Academy of Sciences, Beijing, 100190, China 2.School of Artificial Intelligence, University of Chinese Academy of Sciences, Beijing 100049, China 3.Beijing Unisound Information Technology Co., Ltd, Beijing 100096, 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,He SZ,Liu K,et al. A Unified Shared-Private Network with Denoising for Dialogue State Tracking[J]. JOURNAL OF COMPUTER SCIENCE AND TECHNOLOGY,2021,36(6):1407-1419. |
APA | Liu QB,He SZ,Liu K,Liu SP,&Zhao J.(2021).A Unified Shared-Private Network with Denoising for Dialogue State Tracking.JOURNAL OF COMPUTER SCIENCE AND TECHNOLOGY,36(6),1407-1419. |
MLA | Liu QB,et al."A Unified Shared-Private Network with Denoising for Dialogue State Tracking".JOURNAL OF COMPUTER SCIENCE AND TECHNOLOGY 36.6(2021):1407-1419. |
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 | ||
A Unified Shared-Pri(997KB) | 期刊论文 | 作者接受稿 | 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