Continual Learning of Knowledge Graph Embeddings

被引:35
作者
Daruna, Angel [1 ]
Gupta, Mehul [1 ]
Sridharan, Mohan [2 ]
Chernova, Sonia [1 ]
机构
[1] Georgia Inst Technol, Atlanta, GA 30332 USA
[2] Univ Birmingham, Birmingham, W Midlands, England
基金
美国国家科学基金会;
关键词
Robots; Knowledge engineering; Semantics; Task analysis; Learning systems; Neural networks; Training; Continual learning; representation learning; FRAMEWORK;
D O I
10.1109/LRA.2021.3056071
中图分类号
TP24 [机器人技术];
学科分类号
080202 ; 1405 ;
摘要
In recent years, there has been a resurgence in methods that use distributed (neural) representations to represent and reason about semantic knowledge for robotics applications. However, while robots often observe previously unknown concepts, these representations typically assume that all concepts are known a priori, and incorporating new information requires all concepts to be learned afresh. Our work relaxes this limiting assumption of existing representations and tackles the incremental knowledge graph embedding problem by leveraging the principles of a range of continual learning methods. Through an experimental evaluation with several knowledge graphs and embedding representations, we provide insights about trade-offs for practitioners to match a semantics-driven robotics applications to a suitable continual knowledge graph embedding method.
引用
收藏
页码:1128 / 1135
页数:8
相关论文
共 33 条
[1]   Multimodal estimation and communication of latent semantic knowledge for robust execution of robot instructions [J].
Arkin, Jacob ;
Park, Daehyung ;
Roy, Subhro ;
Walter, Matthew R. ;
Roy, Nicholas ;
Howard, Thomas M. ;
Paul, Rohan .
INTERNATIONAL JOURNAL OF ROBOTICS RESEARCH, 2020, 39 (10-11) :1279-1304
[2]  
Beetz M, 2018, IEEE INT CONF ROBOT, P512
[3]  
Bordes A., 2013, Advances in Neural Information Processing Systems, V26, P2787, DOI DOI 10.5555/2999792.2999923
[4]  
Bowman Samuel R., 2016, SIGNLL, P10, DOI DOI 10.18653/V1/K16-1002
[5]  
Cai L., 2018, P 2018 C N AM CHAPT, V1, P1470, DOI DOI 10.18653/V1/N18-1133
[6]   Situated Bayesian Reasoning Framework for Robots Operating in Diverse Everyday Environments [J].
Chernova, Sonia ;
Chu, Vivian ;
Daruna, Angel ;
Garrison, Haley ;
Hahn, Meera ;
Khante, Priyanka ;
Liu, Weiyu ;
Thomaz, Andrea .
ROBOTICS RESEARCH, 2020, 10 :353-369
[7]   Generative adversarial networks based on Wasserstein distance for knowledge graph embeddings [J].
Dai, Yuanfei ;
Wang, Shiping ;
Chen, Xing ;
Xu, Chaoyang ;
Guo, Wenzhong .
KNOWLEDGE-BASED SYSTEMS, 2020, 190
[8]  
Daruna A, 2019, IEEE INT CONF ROBOT, P9777, DOI [10.1109/ICRA.2019.8794070, 10.1109/icra.2019.8794070]
[9]  
Das Rajarshi, 2018, INT C LEARN REPR
[10]  
Dettmers T, 2018, AAAI CONF ARTIF INTE, P1811