TY - GEN
T1 - Research on the Prospect of Knowledge Graph Completion Based on the Federated Setting
AU - Zhao, Angxiao
AU - Liu, Yunhui
AU - Wei, Songxuan
AU - Long, Yu
AU - Feng, Wenying
AU - Gu, Zhaoquan
N1 - Publisher Copyright:
© 2023 IEEE.
PY - 2023
Y1 - 2023
N2 - Knowledge graph (KG) is a semantic network with graph structure. Knowledge graphs are widely used in industry due to their ease of understanding and coding. Unfortunately, triples in knowledge graphs are usually missing and struggle to cope with complex tasks in existing scenarios. Multi-source knowledge graph is the knowledge graph from different devices, which contribute to the Knowledge Graph Completion (KGC). However, due to the privacy security of graph data from different devices, there are often security risks in sharing real data. Therefore, Knowledge Graph Embedding (KGE) based on federated setting is proposed to solve this problem successfully. In this paper, we conduct extensive experiments based on the federated embedding learning framework FedE, and show the impact of the number of clients and the training data on the accuracy of the graph embedding model, so as to analyze the possibility of using federated Settings in KGC in the future.
AB - Knowledge graph (KG) is a semantic network with graph structure. Knowledge graphs are widely used in industry due to their ease of understanding and coding. Unfortunately, triples in knowledge graphs are usually missing and struggle to cope with complex tasks in existing scenarios. Multi-source knowledge graph is the knowledge graph from different devices, which contribute to the Knowledge Graph Completion (KGC). However, due to the privacy security of graph data from different devices, there are often security risks in sharing real data. Therefore, Knowledge Graph Embedding (KGE) based on federated setting is proposed to solve this problem successfully. In this paper, we conduct extensive experiments based on the federated embedding learning framework FedE, and show the impact of the number of clients and the training data on the accuracy of the graph embedding model, so as to analyze the possibility of using federated Settings in KGC in the future.
KW - Federated learing
KW - Knowledge graph
KW - Knowledge graph completion
KW - Privacy security
UR - https://www.scopus.com/pages/publications/85184351096
U2 - 10.1109/DSC59305.2023.00019
DO - 10.1109/DSC59305.2023.00019
M3 - 会议稿件
AN - SCOPUS:85184351096
T3 - Proceedings - 2023 8th International Conference on Data Science in Cyberspace, DSC 2023
SP - 60
EP - 67
BT - Proceedings - 2023 8th International Conference on Data Science in Cyberspace, DSC 2023
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 8th International Conference on Data Science in Cyberspace, DSC 2023
Y2 - 18 August 2023 through 20 August 2023
ER -