Reasoning for Local Graph Over Knowledge Graph With a Multi-Policy Agent
The reinforcement learning framework for multi-hop relational paths is one of the effective methods for solving knowledge graph incompletion. However, these models are associated with limited performances attributed to delayed rewards and far-fetched search trajectories. To overcome these challenges...
Main Authors: | , , |
---|---|
Format: | Article |
Language: | English |
Published: |
IEEE
2021-01-01
|
Series: | IEEE Access |
Subjects: | |
Online Access: | https://ieeexplore.ieee.org/document/9440923/ |
id |
doaj-3859424523884c2eb82937a04b09374b |
---|---|
record_format |
Article |
spelling |
doaj-3859424523884c2eb82937a04b09374b2021-06-02T23:19:18ZengIEEEIEEE Access2169-35362021-01-019784527846210.1109/ACCESS.2021.30837949440923Reasoning for Local Graph Over Knowledge Graph With a Multi-Policy AgentJie Zhang0https://orcid.org/0000-0002-1135-2031Zhongmin Pei1https://orcid.org/0000-0002-6228-0533Zhangkai Luo2https://orcid.org/0000-0002-1683-5745Science and Technology on Complex Electronic System Simulation Laboratory, Space Engineering University, Beijing, ChinaScience and Technology on Complex Electronic System Simulation Laboratory, Space Engineering University, Beijing, ChinaScience and Technology on Complex Electronic System Simulation Laboratory, Space Engineering University, Beijing, ChinaThe reinforcement learning framework for multi-hop relational paths is one of the effective methods for solving knowledge graph incompletion. However, these models are associated with limited performances attributed to delayed rewards and far-fetched search trajectories. To overcome these challenges, we propose the searching window and multi-policy agent. The searching window provides a large action space, so that the agent can backtrack based on the newly obtained information and establish a local graph instead of a path chain. Based on the searching window, a double long short-term memory (DBL-LSTM) policy network is introduced to encode the local graph and relation sequence, after which the encoding information is used by the agent to select a correct entity to grow the local graph. Furthermore, multi-policy agent separately infers the local graph through three different policy networks, then, all local graphs are integrated into an information-rich local graph. Experiments using the WN18RR dataset revealed that local graph reasoning with searching window had greater rewards than path reasoning, the proposed DBL-LSTM policy network improved all HITS@N(N = 1,3,5,10) compared to prior works, and that the multi-policy agent achieved higher hit rates than single-policy agent.https://ieeexplore.ieee.org/document/9440923/Knowledge graph completionlocal graph reasoningMarkov decision processreinforcement learningsearching window |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Jie Zhang Zhongmin Pei Zhangkai Luo |
spellingShingle |
Jie Zhang Zhongmin Pei Zhangkai Luo Reasoning for Local Graph Over Knowledge Graph With a Multi-Policy Agent IEEE Access Knowledge graph completion local graph reasoning Markov decision process reinforcement learning searching window |
author_facet |
Jie Zhang Zhongmin Pei Zhangkai Luo |
author_sort |
Jie Zhang |
title |
Reasoning for Local Graph Over Knowledge Graph With a Multi-Policy Agent |
title_short |
Reasoning for Local Graph Over Knowledge Graph With a Multi-Policy Agent |
title_full |
Reasoning for Local Graph Over Knowledge Graph With a Multi-Policy Agent |
title_fullStr |
Reasoning for Local Graph Over Knowledge Graph With a Multi-Policy Agent |
title_full_unstemmed |
Reasoning for Local Graph Over Knowledge Graph With a Multi-Policy Agent |
title_sort |
reasoning for local graph over knowledge graph with a multi-policy agent |
publisher |
IEEE |
series |
IEEE Access |
issn |
2169-3536 |
publishDate |
2021-01-01 |
description |
The reinforcement learning framework for multi-hop relational paths is one of the effective methods for solving knowledge graph incompletion. However, these models are associated with limited performances attributed to delayed rewards and far-fetched search trajectories. To overcome these challenges, we propose the searching window and multi-policy agent. The searching window provides a large action space, so that the agent can backtrack based on the newly obtained information and establish a local graph instead of a path chain. Based on the searching window, a double long short-term memory (DBL-LSTM) policy network is introduced to encode the local graph and relation sequence, after which the encoding information is used by the agent to select a correct entity to grow the local graph. Furthermore, multi-policy agent separately infers the local graph through three different policy networks, then, all local graphs are integrated into an information-rich local graph. Experiments using the WN18RR dataset revealed that local graph reasoning with searching window had greater rewards than path reasoning, the proposed DBL-LSTM policy network improved all HITS@N(N = 1,3,5,10) compared to prior works, and that the multi-policy agent achieved higher hit rates than single-policy agent. |
topic |
Knowledge graph completion local graph reasoning Markov decision process reinforcement learning searching window |
url |
https://ieeexplore.ieee.org/document/9440923/ |
work_keys_str_mv |
AT jiezhang reasoningforlocalgraphoverknowledgegraphwithamultipolicyagent AT zhongminpei reasoningforlocalgraphoverknowledgegraphwithamultipolicyagent AT zhangkailuo reasoningforlocalgraphoverknowledgegraphwithamultipolicyagent |
_version_ |
1721399944882946048 |