TY - GEN
T1 - MHPS
T2 - 2024 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2024
AU - Gao, Chen
AU - Qin, Xugong
AU - Zhang, Peng
AU - He, Yongquan
AU - Huang, Xinjian
AU - Zhou, Ming
AU - Zhu, Liehuang
AU - Tan, Qingfeng
N1 - Publisher Copyright:
© 2024 IEEE.
PY - 2024
Y1 - 2024
N2 - Recently, path inference-based knowledge graph reasoning (KGR) methods have attracted great attention due to their good performance and interpretability. However, as the number of hops increases, the search space grows exponentially, making the reward sparse and the process of reasoning difficult. To alleviate this problem, we propose the Multimodality-guided Hierarchical Policy Search (MHPS) for KGR, which introduces multimodal hierarchical guidance to each layer of policies during policy search. On the one hand, multimodal guidance reserves rich information on different dimensions, providing more opportunities to find better paths. On the other hand, this leads to better interaction between the two agents, resulting in more concise guidance for policy stepping. Experimental results on two public datasets demonstrate that the proposed approach outperforms state-of-the-art methods on multi-hop KGR.
AB - Recently, path inference-based knowledge graph reasoning (KGR) methods have attracted great attention due to their good performance and interpretability. However, as the number of hops increases, the search space grows exponentially, making the reward sparse and the process of reasoning difficult. To alleviate this problem, we propose the Multimodality-guided Hierarchical Policy Search (MHPS) for KGR, which introduces multimodal hierarchical guidance to each layer of policies during policy search. On the one hand, multimodal guidance reserves rich information on different dimensions, providing more opportunities to find better paths. On the other hand, this leads to better interaction between the two agents, resulting in more concise guidance for policy stepping. Experimental results on two public datasets demonstrate that the proposed approach outperforms state-of-the-art methods on multi-hop KGR.
KW - Multi-hop knowledge reasoning
KW - Multimodal knowledge graph
UR - http://www.scopus.com/inward/record.url?scp=85195416415&partnerID=8YFLogxK
U2 - 10.1109/ICASSP48485.2024.10447436
DO - 10.1109/ICASSP48485.2024.10447436
M3 - Conference contribution
AN - SCOPUS:85195416415
T3 - ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings
SP - 11096
EP - 11100
BT - 2024 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2024 - Proceedings
PB - Institute of Electrical and Electronics Engineers Inc.
Y2 - 14 April 2024 through 19 April 2024
ER -