Papers › QuatDE: Dynamic Quaternion Embedding for Knowledge Graph Completion

QuatDE: Dynamic Quaternion Embedding for Knowledge Graph Completion

19 May 2021arXiv:2105.09002archive 2025-07-28

Haipeng Gao, Kun Yang, Yuxue Yang, Rufai Yusuf Zakari, Jim Wilson Owusu, Ke Qin

Knowledge graph embedding has been an active research topic for knowledge base completion (KGC), with progressive improvement from the initial TransE, TransH, RotatE et al to the current state-of-the-art QuatE. However, QuatE ignores the multi-faceted nature of the entity and the complexity of the relation, only using rigorous operation on quaternion space to capture the interaction between entitiy pair and relation, leaving opportunities for better knowledge representation which will finally help KGC. In this paper, we propose a novel model, QuatDE, with a dynamic mapping strategy to explicitly capture the variety of relational patterns and separate different semantic information of the entity, using transition vectors to adjust the point position of the entity embedding vectors in the quaternion space via Hamilton product, enhancing the feature interaction capability between elements of the triplet. Experiment results show QuatDE achieves state-of-the-art performance on three well-established knowledge graph completion benchmarks. In particular, the MR evaluation has relatively increased by 26% on WN18 and 15% on WN18RR, which proves the generalization of QuatDE.

PaperPDFCode

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Code

hopkin-ghp/QuatDE officialmentioned in paperpytorch report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Graph EmbeddingKnowledge Base CompletionKnowledge Graph CompletionKnowledge Graph EmbeddingLink Prediction

2 archive task tags without a task page not shown.

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Link Prediction FB15k-237 QuatDE Hits@1 0.268 #17 of 75 Archive leaderboard report
Link Prediction FB15k-237 QuatDE Hits@10 0.563 #17 of 75 Archive leaderboard report
Link Prediction FB15k-237 QuatDE Hits@3 0.40 #17 of 75 Archive leaderboard report
Link Prediction FB15k-237 QuatDE MR 90 #17 of 75 Archive leaderboard report
Link Prediction FB15k-237 QuatDE MRR 0.365 #17 of 75 Archive leaderboard report
Link Prediction WN18 QuatDE Hits@1 0.944 #3 of 37 Archive leaderboard report
Link Prediction WN18 QuatDE Hits@10 0.961 #3 of 37 Archive leaderboard report
Link Prediction WN18 QuatDE Hits@3 0.954 #3 of 37 Archive leaderboard report
Link Prediction WN18 QuatDE MR 120 #3 of 37 Archive leaderboard report
Link Prediction WN18 QuatDE MRR 0.95 #3 of 37 Archive leaderboard report
Link Prediction WN18RR QuatDE Hits@1 0.438 #20 of 75 Archive leaderboard report
Link Prediction WN18RR QuatDE Hits@10 0.586 #20 of 75 Archive leaderboard report
Link Prediction WN18RR QuatDE Hits@3 0.509 #20 of 75 Archive leaderboard report
Link Prediction WN18RR QuatDE MR 1977 #20 of 75 Archive leaderboard report
Link Prediction WN18RR QuatDE MRR 0.489 #20 of 75 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Methods

RotatESelf-Adversarial Negative SamplingTransE

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections