Knowledge Commons of Institute of Automation,CAS
Information bottleneck based knowledge selection for commonsense reasoning | |
Zhao Yang1,2; Yuanzhe Zhang1,2; Pengfei Cao1,2; Cao Liu3; Jiansong Chen3; Jun Zhao1,2; Kang Liu1,2,4 | |
发表期刊 | Information Sciences |
ISSN | 0020-0255 |
2024 | |
卷号 | 660页码:120134 |
通讯作者 | Liu, Kang(liucao@meituan.com) |
摘要 | KG-augmented models usually endow existing models with external knowledge graphs, which achieve promising performance in various knowledge-intensive tasks, such as commonsense reasoning. Existing methods mainly first exploited heuristic ways for retrieving the relevant knowledge subgraphs according to the input, and then utilized some effective encoders, such as GNNs, to encode the symbolic knowledge into the neural reasoning networks. However, whether the whole retrieved knowledge subgraphs are really relevant or useful for the reasoning process was seldom considered. Actually, according to our observations and analysis, most retrieved knowledge is noisy and useless to the reasoning models, which would hurt the final performance. To remedy this, this paper proposes information bottleneck based knowledge selection (IBKS), which is able to select useful knowledge from the retrieved knowledge subgraph. Expectedly, the selected knowledge could better improve the commonsense reasoning ability of the model. Moreover, IBKS is model-agnostic and could be plugged into any existing KG-augmented model. Extensive experimental results show that IBKS could effectively improve commonsense reasoning performance. |
关键词 | Commonsense reasoning Knowledge selection Information bottleneck KG-augmented model |
DOI | https://doi.org/10.1016/j.ins.2024.120134 |
收录类别 | SCI |
语种 | 英语 |
资助项目 | National Key R&D Program of China[2022ZD0160503] ; National Natural Science Foundation of China[62276264] ; Strategic Priority Research Program of Chinese Academy of Sciences[XDA27020100] ; Youth Innovation Promotion Association CAS Project[202202AD080004] ; Yunnan Provincial Major Science and Technology Special Plan Project Project[202202AD080004] ; Meituan |
项目资助者 | National Key R&D Program of China ; National Natural Science Foundation of China ; Strategic Priority Research Program of Chinese Academy of Sciences ; Youth Innovation Promotion Association CAS Project ; Yunnan Provincial Major Science and Technology Special Plan Project Project ; Meituan |
WOS研究方向 | Computer Science |
WOS类目 | Computer Science, Information Systems |
WOS记录号 | WOS:001164241400001 |
出版者 | ELSEVIER SCIENCE INC |
七大方向——子方向分类 | 自然语言处理 |
国重实验室规划方向分类 | 语音语言处理 |
是否有论文关联数据集需要存交 | 否 |
引用统计 | |
文献类型 | 期刊论文 |
条目标识符 | http://ir.ia.ac.cn/handle/173211/56721 |
专题 | 复杂系统认知与决策实验室 |
通讯作者 | Kang Liu |
作者单位 | 1.School of Artificial Intelligence, University of Chinese Academy of Sciences, China 2.The Laboratory of Cognition and Decision Intelligence for Complex Systems, Institute of Automation, Chinese Academy of Sciences, China 3.Meituan, Beijing, China 4.Shanghai Artificial Intelligence Laboratory, China |
第一作者单位 | 中国科学院自动化研究所 |
通讯作者单位 | 中国科学院自动化研究所 |
推荐引用方式 GB/T 7714 | Zhao Yang,Yuanzhe Zhang,Pengfei Cao,et al. Information bottleneck based knowledge selection for commonsense reasoning[J]. Information Sciences,2024,660:120134. |
APA | Zhao Yang.,Yuanzhe Zhang.,Pengfei Cao.,Cao Liu.,Jiansong Chen.,...&Kang Liu.(2024).Information bottleneck based knowledge selection for commonsense reasoning.Information Sciences,660,120134. |
MLA | Zhao Yang,et al."Information bottleneck based knowledge selection for commonsense reasoning".Information Sciences 660(2024):120134. |
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 | ||
杨朝-Information Scien(1069KB) | 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