Papers › TuckER: Tensor Factorization for Knowledge Graph Completion

TuckER: Tensor Factorization for Knowledge Graph Completion

28 Jan 2019IJCNLP 2019 11arXiv:1901.09590archive 2025-07-28

Ivana Balažević, Carl Allen, Timothy M. Hospedales

Knowledge graphs are structured representations of real world facts. However, they typically contain only a small subset of all possible facts. Link prediction is a task of inferring missing facts based on existing ones. We propose TuckER, a relatively straightforward but powerful linear model based on Tucker decomposition of the binary tensor representation of knowledge graph triples. TuckER outperforms previous state-of-the-art models across standard link prediction datasets, acting as a strong baseline for more elaborate models. We show that TuckER is a fully expressive model, derive sufficient bounds on its embedding dimensionalities and demonstrate that several previously introduced linear models can be viewed as special cases of TuckER.

PaperPDFConference PDFCode

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

Code

ibalazevic/TuckER officialmentioned in papermentioned on GitHubpytorchMIT report
Sujit-O/pykg2vec mentioned on GitHubtfMIT report
adaruna3/explainable-kge mentioned on GitHubpytorch report
allenai/kb mentioned on GitHubpytorchApache-2.0 report
luffycodes/neptune mentioned on GitHubpytorch 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

Knowledge Graph CompletionKnowledge GraphsLink PredictionPrediction

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Link Prediction FB15k TuckER Hits@1 0.741 #9 of 23 Archive leaderboard report
Link Prediction FB15k TuckER Hits@10 0.892 #9 of 23 Archive leaderboard report
Link Prediction FB15k TuckER Hits@3 0.833 #9 of 23 Archive leaderboard report
Link Prediction FB15k TuckER MRR 0.795 #9 of 23 Archive leaderboard report
Link Prediction FB15k-237 TuckER Hits@1 0.266 #22 of 75 Archive leaderboard report
Link Prediction FB15k-237 TuckER Hits@10 0.544 #22 of 75 Archive leaderboard report
Link Prediction FB15k-237 TuckER Hits@3 0.394 #22 of 75 Archive leaderboard report
Link Prediction FB15k-237 TuckER MRR 0.358 #22 of 75 Archive leaderboard report
Link Prediction WN18 TuckER Hits@1 0.949 #10 of 37 Archive leaderboard report
Link Prediction WN18 TuckER Hits@10 0.958 #10 of 37 Archive leaderboard report
Link Prediction WN18 TuckER Hits@3 0.955 #10 of 37 Archive leaderboard report
Link Prediction WN18 TuckER MRR 0.953 #10 of 37 Archive leaderboard report
Link Prediction WN18RR TuckER Hits@1 0.443 #60 of 75 Archive leaderboard report
Link Prediction WN18RR TuckER Hits@10 0.526 #60 of 75 Archive leaderboard report
Link Prediction WN18RR TuckER Hits@3 0.482 #60 of 75 Archive leaderboard report
Link Prediction WN18RR TuckER MRR 0.470 #60 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

Introduced by this paper: TuckER

TuckER

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