The DOI matches its Crossref record; 23 of the 23 references it cites were found in a registry or an index; 5 of the 14 sources its claims cite were checked for support.
TuckER: Tensor Factorization for Knowledge Graph Completion
Open accessAbstract
Ivana Balazevic, Carl Allen, Timothy Hospedales. Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP). 2019.
Overview
AI snapshot
Written from the paper’s full text.
TL;DR
TuckER is a linear link-prediction model based on Tucker decomposition of the binary knowledge-graph tensor, using a shared core tensor and entity/relation embeddings. It achieves state-of-the-art results on standard datasets, is fully expressive, and subsumes several prior linear models as special cases.
Key findings
- TuckER reaches state-of-the-art link prediction results on WN18RR (MRR .470, Hits@10 .526, Hits@3 .482, Hits@1 .443) and FB15k-237 (MRR .358, Hits@10 .544, Hits@3 .394, Hits@1 .266), and on WN18 (MRR .953, Hits@10 .958, Hits@3 .955, Hits@1 .949) and FB15k (MRR .795, Hits@10 .892, Hits@3 .833, Hits@1 .741), apart from Hits@10 on WN18 where R-GCN does better.
- TuckER is fully expressive: Theorem 1 gives a sufficient bound of entity embedding dimensionality \(d_e = n_e\) and relation embedding dimensionality \(d_r = n_r\).
- RESCAL, DistMult, ComplEx, and SimplE can be viewed as special cases of TuckER.
- Parameter sharing through the core tensor helps: on WN18RR TuckER uses ∼9.4M parameters versus ∼16.4M for ComplEx and SimplE at \(d_e=200, d_r=30\); at embedding size 20 (∼300k parameters), TuckER is almost as good as ComplEx and SimplE at embedding size 200 (∼6M parameters), and on FB15k it improves Hits@1 by +14% over ComplEx and +8% over SimplE.
Methods
- TuckER factorizes the third-order binary tensor of triples with Tucker decomposition, scoring \(\phi(e_s, r, e_o) = W \times_1 e_s \times_2 w_r \times_3 e_o\), where \(W\) is a core tensor and subject/object entity embeddings are shared; a logistic sigmoid converts scores to probabilities.
- Training uses reciprocal relations, 1-N scoring, Bernoulli negative log-likelihood, Adam with batch size 128, and random-search hyperparameters; evaluation uses the filtered setting with MRR and Hits@k on FB15k, FB15k-237, WN18, and WN18RR.
- Theoretical analysis proves full expressiveness, derives the dimensionality bound, and shows the relationship between TuckER and prior linear tensor-factorization models.
Limitations
- The paper states future work might include exploring how to incorporate background knowledge on individual relation properties; it does not present a dedicated limitations section.
- Evident from the text: evaluation is limited to four standard link-prediction datasets and MRR/Hits@k metrics, with no broader task or computational-efficiency analysis beyond parameter counts.
- Evident from the text: the full-expressiveness bound \(d_e = n_e, d_r = n_r\) is theoretical and may be much larger than practical embedding sizes; the authors expect smaller dimensionalities to suffice in practice.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
22 Own interpretation · 9 Own result · 6 Sources not found · 3 Partly supported · 3 Own method · 3 No basis given · 2 Supported · 1 No full text of the sources · accuracy not yet measured
Every source the paper cites for this claim was checked and supports it.
In the paper
“The authors show that convolution is a way of introducing sparsity and parameter tying and that HypER can be understood in terms of tensor factorization up to a non-linearity, thus placing HypER closer to the well established family of factorization models.”
§ TuckER: Tensor Factorization for Knowledge Graph Completion > 1 Introduction
What the sources say
- Supportedcites Balažević et al., 2019 source
Section 4.2 (p. 6) describes HypER as tensor multiplication with a constrained core tensor and notes that ReLU is applied before the final dot product. The Related Work section (p. 3) explicitly identifies that ReLU non-linearity as the key difference from existing factorization approaches and describes convolution as multiplication by a sparse tensor with tied weights. (LLM, conf=0.94)
boostedcorrelational- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“They further prove other models are not fully expressive: DistMult, because it cannot model asymmetric relations; and transitive models such as TransE (Bordes et al., 2013) and its variants FTransE (Feng et al., 2016) and STransE (Nguyen et al., 2016), because of certain contradictions that they impose between different relation types.”
§ TuckER: Tensor Factorization for Knowledge Graph Completion > 5.1 Full Expressiveness and Embedding Dimensionality
What the sources say
- Not matchedcites Bordes et al., 2013
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
- Not matchedcites Feng et al., 2016
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
- Supportedcites Nguyen et al., 2016 source
The citation immediately following “STransE” identifies the model, rather than attributing the citing authors’ expressiveness proof to Nguyen et al. The source’s model description (SE section; page unavailable) explicitly introduces STransE. It does not establish the claimed contradictions, but that proof is framed as the citing authors’ own. (LLM, conf=0.91)
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Not matched
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“To our knowledge, ComplEx-N3 (Lacroix et al., 2018) is the only other linear link prediction model that benefits from multitask learning.”
§ TuckER: Tensor Factorization for Knowledge Graph Completion > 6.3 Link Prediction Results
What the sources say
- Partly supportedcites Lacroix et al., 2018 sourceExperimental — accuracy not yet measured
The Experimental Setup (p. 7) identifies ComplEx-N3, and the Introduction (pp. 1–2) reports performance gains for ComplEx from modeling inverse predicates and, on harder datasets, from N3 regularization. Our copy does not establish that ComplEx-N3 benefits from multitask learning or that it is the only other linear link-prediction model to do so. (LLM, conf=0.78)
hedged- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Partly supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“There, rank regularization of the embedding matrices is used to encourage a lowrank factorization, thus forcing parameter sharing between relations.”
§ TuckER: Tensor Factorization for Knowledge Graph Completion > 6.3 Link Prediction Results
What the sources say
- Partly supportedcites Lacroix et al., 2018 sourceExperimental — accuracy not yet measured
Section 4 (p. 4) confirms that the paper proposes nuclear 3-norm regularization for tensor-factorization models, including ComplEx. But it does not say that this regularizer forces parameter sharing between relations; Section 2.2 (p. 2) instead describes ComplEx’s parameter sharing as occurring between the subject and object modes. (LLM, conf=0.84)
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Partly supported
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“TuckER outperforms previous state-of-the-art models across standard link prediction datasets, acting as a strong baseline for more elaborate models.”
- Also stated in: § 1 Introduction · plain
“proposing TuckER, a new linear model for link prediction on knowledge graphs, that is simple, expressive and achieves state-of-theart results across all standard datasets”
§ TuckER: Tensor Factorization for Knowledge Graph Completion > Abstract
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- The authors state this as an assumption: Subject and object entity embedding matrices are equivalent, i.e. no distinction is made between the embeddings of an entity depending on whether it appears as a subject or as an object.
- Also stated in: § 1 Introduction · plain
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“This is supported by the fact that the margin by which TuckER outperforms other linear models is notably increased on datasets with a large number of relations.”
§ TuckER: Tensor Factorization for Knowledge Graph Completion > 6.3 Link Prediction Results
plaincorrelational- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“Even with fewer parameters than ComplEx and SimplE at d e = 200 and d r = 30 on WN18RR (∼9.4 vs ∼16.4 million), TuckER consistently obtains better results than any of those models.”
§ TuckER: Tensor Factorization for Knowledge Graph Completion > 6.3 Link Prediction Results
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“We can see that the difference between the MRRs of ComplEx, SimplE and TuckER is approximately constant for embedding sizes 100 and 200.”
§ TuckER: Tensor Factorization for Knowledge Graph Completion > 6.4 Influence of Parameter Sharing
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“However, running their model with equivalent parameter number to TuckER shows comparable performance, supporting our belief that the two models both attain the benefits of multi-task learning, although by different means.”
§ TuckER: Tensor Factorization for Knowledge Graph Completion > 6.3 Link Prediction Results
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“It is important to note that at embedding dimensionalities 20, 50 and 100, TuckER has fewer parameters than Com-plEx and SimplE (e.g.ComplEx and SimplE have ∼3 million and TuckER has ∼2.5 million parameters for embedding dimensionality 100).”
§ TuckER: Tensor Factorization for Knowledge Graph Completion > 6.4 Influence of Parameter Sharing
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Probably not a substantial claim. 9
- It may state two separate claims as one. 9
Every source the paper cites for this claim was checked and supports it.
In the paper
“As shown in (Trouillon et al., 2017), ComplEx is fully expressive with the embedding dimensionality bound d e = d r = n e • n r .”
§ TuckER: Tensor Factorization for Knowledge Graph Completion > 5.1 Full Expressiveness and Embedding Dimensionality
What the sources say
- Supportedcites Trouillon et al., 2017 source
Theorem 4 in §3.2 (p. 14) shows that ComplEx can represent arbitrary families of relation matrices using a shared entity embedding matrix and diagonal relation matrices of dimension nm, where n is the number of entities and m the number of relations. It states that the decomposition rank is at most nm, matching the claimed dimensionality bound for entity and relation embeddings. (LLM, conf=0.92)
boosted- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
- Supported
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“However, for lower embedding sizes, the dif-ference between MRRs increases by 0.7% for embedding size 50 and by 4.2% for embedding size 20 for ComplEx and by 3% for embedding size 50 and by 9.9% for embedding size 20 for Sim-plE.”
§ TuckER: Tensor Factorization for Knowledge Graph Completion > 6.4 Influence of Parameter Sharing
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Results achieved by TuckER are not only better than those of other linear models, such as DistMult, ComplEx and SimplE, but also better than the results of many more complex deep neural network and reinforcement learning architectures, e.g.R-GCN, MINERVA, ConvE and HypER, demonstrating the expressive power of linear models and supporting our claim that simple linear models should serve as a”
§ TuckER: Tensor Factorization for Knowledge Graph Completion > 6.3 Link Prediction Results
boosted- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Overall, TuckER outperforms previous state-of-the-art models on all metrics across all datasets (apart from hits@10 on WN18 where a non-linear model, R-GCN, does better).”
§ TuckER: Tensor Factorization for Knowledge Graph Completion > 6.3 Link Prediction Results
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“For example, improvement on FB15k is +14% over ComplEx and +8% over SimplE on the toughest hits@1 metric.”
§ TuckER: Tensor Factorization for Knowledge Graph Completion > 6.3 Link Prediction Results
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“It is trivial to show that both RESCAL and DistMult belong to the family of bilinear models.”
§ TuckER: Tensor Factorization for Knowledge Graph Completion > 5.2 Relation to Previous Linear Models
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
In the paper
“It is interesting to note that the TuckER interpretation of the DistMult scoring function, given that matrices A and C are identical, can alternatively be interpreted as a special case of CP decomposition (Hitchcock, 1927), since Tucker decomposition with a superdiagonal core tensor is equivalent to CP decomposition.”
§ TuckER: Tensor Factorization for Knowledge Graph Completion > 5.2 Relation to Previous Linear Models
What the sources say
- Not checked — no full text of the sourcecites Hitchcock, 1927 source
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“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.”
- Also stated in: § 1 Introduction · plain
“Finally, we show that several previous stateof-the-art linear models, RESCAL (Nickel et al., 2011), DistMult (Yang et al., 2015), ComplEx (Trouillon et al., 2016) and SimplE (Kazemi and Poole, 2018), are special cases of TuckER.”
- Also stated in: § 1 Introduction · plain
“showing how TuckER subsumes several previously proposed tensor factorization approaches to link prediction.”
§ TuckER: Tensor Factorization for Knowledge Graph Completion > Abstract
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
- The authors state this as an assumption: Subject and object entity embedding matrices are equivalent, i.e. no distinction is made between the embeddings of an entity depending on whether it appears as a subject or as an object.
- Also stated in: § 1 Introduction · plain
In the paper
“Similarly to Com-plEx, Kazemi and Poole (2018) show that SimplE is fully expressive with entity and relation embeddings of size d e = d r = min(n e • n r , γ + 1), where γ represents the number of true facts.”
§ TuckER: Tensor Factorization for Knowledge Graph Completion > 5.1 Full Expressiveness and Embedding Dimensionality
What the sources say
- Not matchedcites Kazemi and Poole (2018)
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
- Not matched
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“So far, there have been two possible ways in which linear link prediction models introduce asymmetry into factorization of the binary tensor of triples: • distinct (although possibly related) embeddings for subject and object entities and a diagonal matrix (or equivalently a vector) for each relation, as is the case with models such as ComplEx and SimplE; or • equivalent subject and object entity”
§ TuckER: Tensor Factorization for Knowledge Graph Completion > 5.3 Representing Asymmetric Relations
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. It may state two separate claims as one.
What the paper did — it cites no source for this, so there is none to check. The quote shows it is in the paper.
In the paper
“We use evaluation metrics standard across the link prediction literature: mean reciprocal rank (MRR) and hits@k, k ∈ {1, 3, 10}.”
§ TuckER: Tensor Factorization for Knowledge Graph Completion > 6.2 Implementation and Experiments
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“TuckER overcomes this by representing relations as vectors w r , which makes the parameter number grow linearly with the number of relations, while still keeping the desirable property of allowing relations to be asymmetric”
§ TuckER: Tensor Factorization for Knowledge Graph Completion > 5.3 Representing Asymmetric Relations
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
- Rests on an assumption: The full rank matrix approach for relations often leads to overfitting, especially for relations with a small number of training triples.
What the paper did — it cites no source for this, so there is none to check. The quote shows it is in the paper.
In the paper
“To test this, we re-implement ComplEx and SimplE with reciprocal relations, 1-N scoring, batch normalization and dropout for fair comparison, perform random search to choose best hyper-parameters”
§ TuckER: Tensor Factorization for Knowledge Graph Completion > 6.4 Influence of Parameter Sharing
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“TuckER achieves state-of-the-art results on standard link prediction datasets, in part due to its ability to perform multi-task learning across relations.”
§ TuckER: Tensor Factorization for Knowledge Graph Completion > 7 Conclusion
plainconditional- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
In the paper
“We use the filtered setting (Bordes et al., 2013), i.e. all known true triples are removed from the candidate set except for the current test triple.”
§ TuckER: Tensor Factorization for Knowledge Graph Completion > 6.2 Implementation and Experiments
What the sources say
- Not matchedcites Bordes et al., 2013
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim.
- Not matched
What the paper did — it cites no source for this, so there is none to check. The quote shows it is in the paper.
In the paper
“We choose all hyper-parameters by random search based on validation set performance.”
§ TuckER: Tensor Factorization for Knowledge Graph Completion > 6.2 Implementation and Experiments
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
The paper relies on this without support.
Knowledge graphs typically contain only a small subset of all possible facts.
“However, they typically contain only a small subset of all possible facts.”
“However, far from all available information is currently stored in existing knowledge graphs”
The paper relies on this without support.
Manually adding new information to knowledge graphs is costly.
“and manually adding new information is costly”
The authors state this as an assumption.
Subject and object entity embedding matrices are equivalent, i.e. no distinction is made between the embeddings of an entity depending on whether it appears as a subject or as an object.
“Subject and object entity embedding matrices are assumed equivalent, i.e. we make no distinction between the embeddings of an entity depending on whether it appears as a subject or as an object in a particular triple.”
The paper relies on this without support.
The full rank matrix approach for relations often leads to overfitting, especially for relations with a small number of training triples.
“which often leads to overfitting, especially for relations with a small number of training triples.”
The authors state this as an assumption.
TuckER should need a lower number of parameters for obtaining good results than ComplEx or SimplE.
“The ability of knowledge sharing through the core tensor suggests that TuckER should need a lower number of parameters for obtaining good results than ComplEx or SimplE.”
“which supports our initial assumption.”
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
Each contradiction between two passages listed here was confirmed by an independent check.
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ TuckER: Tensor Factorization for Knowledge Graph Completion > Model
“RESCAL (Nickel et al., 2011) e s W r e o W r ∈ R de 2 O(n e d e + n r d 2 r )”
§ TuckER: Tensor Factorization for Knowledge Graph Completion > 5.2 Relation to Previous Linear Models
“the relation matrices W r ∈ R de×de are the d e × d e slices of the core tensor Z.”
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ TuckER: Tensor Factorization for Knowledge Graph Completion > Model
“W ∈ R de×de×dr is the core tensor of a Tucker decomposition.”
§ TuckER: Tensor Factorization for Knowledge Graph Completion > 4 Tucker Decomposition for Link Prediction
“w r ∈ R dr the rows of R representing the relation embedding vector and W ∈ R de×dr×de is the core tensor.”
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ TuckER: Tensor Factorization for Knowledge Graph Completion > 6.3 Link Prediction Results
“Overall, TuckER outperforms previous state-of-the-art models on all metrics across all datasets (apart from hits@10 on WN18 where a non-linear model, R-GCN, does better).”
§ TuckER: Tensor Factorization for Knowledge Graph Completion > Additional information
“| HypER (Balažević et al., 2019) | no | .951 | 958 | .955 | .947 | .790 | .885 | .829 | .734 | | SimplE (Kazemi and Poole, 2018) | yes | .942 | .947 | .944 | .939 | .727 | .838 | .773 | .660 | | TuckER (ours) | yes | .953 | .958 | .955 | .949 | .795 | .892 | .833 | .741 |”
Citations
Cites
The publisher hasn’t deposited this paper’s reference list.
Research works printed but not linked: 6
- Antoine Bordes, Nicolas Usunier, Alberto Garcia- Duran, Jason Weston, and Oksana Yakhnenko. 2013. Translating Embeddings for Modeling Multi- relational Data. In Advances in Neural Information Processing Systems.
- Rajarshi Das, Shehzaad Dhuliawala, Manzil Zaheer, Luke Vilnis, Ishan Durugkar, Akshay Krishna- murthy, Alex Smola, and Andrew McCallum. 2018. Go for a Walk and Arrive at the Answer: Reason- ing over Paths in Knowledge Bases Using Rein- forcement Learning. In International Conference on Learning Representations.
- Seyed Mehran Kazemi and David Poole. 2018. Sim- plE Embedding for Link Prediction in Knowledge Graphs. In Advances in Neural Information Pro- cessing Systems.
- Maximilian Nickel, Volker Tresp, and Hans-Peter Kriegel. 2011. A Three-Way Model for Collective Learning on Multi-Relational Data. In International Conference on Machine Learning.
- Adam Paszke, Sam Gross, Soumith Chintala, Gre- gory Chanan, Edward Yang, Zachary DeVito, Zem- ing Lin, Alban Desmaison, Luca Antiga, and Adam Lerer. 2017. Automatic Differentiation in PyTorch. In NIPS-W.
- Bishan Yang, Wen-tau Yih, Xiaodong He, Jianfeng Gao, and Li Deng. 2015. Embedding Entities and Relations for Learning and Inference in Knowledge Bases. In International Conference on Learning Representations.
Cited by
Sorted by citations
Items 301–400 of 699
$$\mu \text {KG}$$: A Library for Multi-source Knowledge Graph Embeddings and Applications
Luo, X., Sun, Z., & Hu, W. P. (2022) · Lecture notes in computer science · doi:10.1007/978-3-031-19433-7_35 · Cited by 5
Learning temporal granularity with quadruplet networks for temporal knowledge graph completion
Geng, R., & Luo, C. (2025) · Scientific Reports · doi:10.1038/s41598-025-00446-z · Cited by 4
MobiFM: A Foundation Model for Mobile Data Forecasting
Chai, H., Qi, X., Ma, Y., Wang, Z., Yue, L., & Li, Y. (2025) · IEEE Journal on Selected Areas in Communications · doi:10.1109/jsac.2025.3642851 · Cited by 4
“To obtain effective representations of entities and relations in the SKG, we draw on the vector decomposition strategy in the TuckER algorithm [39], which is used to extract representation vectors related to mobile data while preserving the structured information of the original SKG.”
Luo, S., Lu, X. X., Zhao, Q., & Rao, W. (2026) · ACM International Conference on Web Search and Data Mining (WSDM) · doi:10.1145/3773966.3777934 · Cited by 4
Zhou, D., Tong, H., Wang, L., Liu, S., Xiong, X., Gan, Z., Griffier, R., Hejblum, B. P., Liu, Y., Hong, C., et al. (2026) · Nature Communications · doi:10.1038/s41467-026-71152-1 · Cited by 15
Multi-hop reasoning over sparse knowledge graphs with deep reinforcement learning
Liu, Y., Fu, L.-M., Fu, Y.-K., Wu, T., & Bai, H. (2025) · Expert Systems with Applications · doi:10.1016/j.eswa.2025.128289 · Cited by 4
NePTuNe: Neural Powered Tucker Networkfor Knowledge Graph Completion
Sonkar, S., Katiyar, A., & Baraniuk, R. G. (2021) · International Joint Conference on Knowledge Graphs (IJCKG) · doi:10.1145/3502223.3502249 · Cited by 5
“Tensor factorization methods represent the knowledge graph as a 3-mode binary tensor (Nickel et al., 2011; Yang et al., 2015; Trouillon et al., 2017; Balazevic et al., 2019c).”
BackgroundMethodInfluentialStudy and analysis of various link predictions in knowledge graph: A challenging overview
Khobragade, A., & Ghumbre, S. U. (2022) · Intelligent Decision Technologies · doi:10.3233/idt-210103 · Cited by 4
“Balaževic et al. [26] developed a Tensor factorization for the knowledge graph completion, called the TuckER.”
HyperKGQA: Question Answering over Knowledge Graphs using Hyperbolic Representation Learning
Madjid, N. A., Khatib, O. E., Gao, S., & Difallah, D. E. (2022) · Proceedings · doi:10.1109/icdm54844.2022.00041 · Cited by 5
LitCQD: Multi-hop Reasoning in Incomplete Knowledge Graphs with Numeric Literals
Demir, C., Wiebesiek, M., Lu, R., et al. (2023) · Lecture notes in computer science · doi:10.1007/978-3-031-43418-1_37 · Cited by 4
Multi-Aspect Enhanced Convolutional Neural Networks for Knowledge Graph Completion
Zhang, F., Qiu, P., Shen, T., & Cheng, J. (2023) · Frontiers in artificial intelligence and applications · doi:10.3233/faia230615 · Cited by 4
EwLoss: A Exponential Weighted Loss Function for Knowledge Graph Embedding Models
Shen, Q., Zhang, H., & Liao, C. (2023) · IEEE Access · doi:10.1109/access.2023.3322204 · Cited by 4
“In many cases, the binary cross-entropy loss is widely used in models like TuckER [36], MuRP [37], ConvE [38], SACN [39], HypER [40], R-GCN [41], ConvR [42], and MKBE [43].”
MethodSchema First! Learn Versatile Knowledge Graph Embeddings by Capturing Semantics with MASCHInE
Hubert, N., Paulheim, H., Monnin, P., et al. (2023) · ACM International Conference on Knowledge Capture (K-CAP) · doi:10.1145/3587259.3627550 · Cited by 4
“Most of the KGEMs subsequently proposed aim at addressing its limitations and improve the representation expressiveness of KGEs [21], such as DistMult [24], ComplEx [20], ConvE [8], and TuckER [1].”
BackgroundMethodInfluentialDual-Safety Knowledge Graph Completion for Process Industry
Wang, L., Li, H., Wang, W., Xin, G., & Wei, Y. (2024) · Electronics · doi:10.3390/electronics13010214 · Cited by 4
MHRE: Multivariate link prediction method for medical hyper-relational facts
Wang, W., Zhang, X., Zhang, J., Cai, W., Zhao, H., & Zhang, X. (2024) · Applied Intelligence · doi:10.1007/s10489-023-05248-2 · Cited by 4
“GETD [21] extends TuckER [22] and utilizes tensor decomposition on hyper-relational facts for link prediction.”
Symbolic Knowledge Reasoning on Hyper-Relational Knowledge Graphs
Wang, Z., Li, L., & Zeng, D. D. (2024) · IEEE Transactions on Big Data · doi:10.1109/tbdata.2024.3423670 · Cited by 4
“Tensor factorization methods on KGs include DistMult [9], TuckER [10], SimplE [11], etc. m-DistMult [1] extends DistMult to higher orders and applies it to hyper-relational facts.”
Shi, L., Liu, W., Wu, Y., Dai, C., Ji, Z., & Ганчев, И. (2024) · Mathematics · doi:10.3390/math12182821 · Cited by 4
Confidence-aware Self-Semantic Distillation on Knowledge Graph Embedding
Liu, Y., Chen, J., Chen, D., Zhou, Z., Feng, Y., & Wang, C. (2024) · ACM International Conference on Information and Knowledge Management (CIKM) · doi:10.1145/3627673.3679683 · Cited by 4
“To investigate the model performance in the low-dimension space, we set the embedding dimensions to 25 and 100.”
MethodSurvey on Computational Applications of Tensor-Network Simulations
García, M. D., & Romero, A. M. (2024) · IEEE Access · doi:10.1109/access.2024.3519676 · Cited by 13
Automating Data Science Pipelines with Tensor Completion
Pakala, S., Graw, B., Ahn, D., Dinh, T., Mahin, M. T., Tsotras, V. J., Chen, J., & Papalexakis, E. E. (2024) · IEEE International Conference on Big Data (BigData) · doi:10.1109/bigdata62323.2024.10825934 · Cited by 4
Predicting affinity ties in a surname network
Mendoza, M., & Bro, N. (2021) · PLoS ONE · doi:10.1371/journal.pone.0256603 · Cited by 4
Modeling IsA Relations via Box Structure for Knowledge Graph Embedding
Dong, Y., Wang, L., Ji, X., & Liu, K. (2022) · Lecture notes in computer science · doi:10.1007/978-3-031-05936-0_24 · Cited by 4
ProtoE: Enhancing Knowledge Graph Completion Models with Unsupervised Type Representation Learning
Lu, Y., & Ichise, R. (2022) · Information · doi:10.3390/info13080354 · Cited by 4
Ni, R., Shibata, H., & Takama, Y. (2022) · IEEE International Conference on Big Data (BigData) · doi:10.1109/bigdata55660.2022.10020261 · Cited by 4
“TuckER [1] is used as the base model for our experiments, and to examine the effectiveness of entity type information, two base models with different hyperparameter settings are tested: TuckER has the same entity dimensionality as EETCRL-TuckER and the entity dimensionality of TuckER1 is equal to the sum of the entity…”
MethodInfluentialSparseMult: A Tensor Decomposition model based on Sparse Relation Matrix
Xie, Z., Zhu, R., Zhang, M., et al. (2022) · IEEE/WIC/ACM International Joint Conference on Web Intelligence and Intelligent Agent Technology (WI-IAT) · doi:10.1109/wi-iat55865.2022.00124 · Cited by 4
“Experimental results show that our SparseMult model not only outperforms Rescal and DistMult models but also achieve better performance than state-of-the-art methods, such as RotatE [8] and TuckER [9].”
MethodInfluentialOpenFact: Factuality Enhanced Open Knowledge Extraction
Song, L., Wang, A., Pan, X., Zhang, H., Yu, D., Jin, L., Mi, H., Su, J., Zhang, Y., & Yu, D. (2023) · Transactions of the Association for Computational Linguistics · doi:10.1162/tacl_a_00569 · Cited by 4
“OpenFact for KBC We do not change the model structure but propose to incorporate one relevant OpenFact triplet to enhance the corresponding candidate KG triple, so that the KG-BERT model in the pipeline system can leverage richer information to make more accurate ranking decisions.”
BackgroundInfluentialHRotatE: Hybrid Relational Rotation Embedding for Knowledge Graph
Shah, A., Molokwu, B. C., & Kobti, Z. (2021) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn52387.2021.9533416 · Cited by 4
“They are not transparent and poorly understood in comparison to other approaches [22].”
BackgroundMethodResultInfluentialA Novel Embedding Model for Knowledge Graph Completion Based on Quaternion
Gao, H., Yang, K., Yang, Y., & Qin, K. (2021) · IEEE International Conference on Information, Communication and Networks (ICICN) · doi:10.1109/icicn52636.2021.9673862 · Cited by 4
What Is Learned in Knowledge Graph Embeddings?
Douglas, M. R., Simkin, M., Ben‐Eliezer, O., et al. (2022) · Studies in computational intelligence · doi:10.1007/978-3-030-93413-2_49 · Cited by 4
“If we take the latent variables to be class probabilities, these look rather similar to embedding models such as TuckER [5].”
MethodSemantic and Relational-Based Graph Neural Network for Knowledge Graph CompletionPreprint
Li, X., Tian, Y., & Ji, S. (2023) · SSRN Electronic Journal · doi:10.2139/ssrn.4455564 · Cited by 4
MicroKGCL: A Knowledge Graph for Root Cause Localization of Feedback Issues in Microservices
Yao, X.-W., Lu, Q.-C., Li, Q., Liu, L.-L., & Zhu, Z.-C. (2023) · 2023 IEEE 23rd International Conference on Software Quality, Reliability, and Security (QRS) · doi:10.1109/qrs60937.2023.00067 · Cited by 3
Zhou, G., Haihong, E., Kuang, Z., Ling, T., Yao, T., & Song, M. (2024) · IEEE Journal of Biomedical and Health Informatics · doi:10.1109/jbhi.2024.3503061 · Cited by 4
Interleaving and cross-attention presents efficient knowledge graph embedding
Wu, J., Laili, Y., Qi, C., Li, Z., & Ren, L. (2025) · Expert Systems with Applications · doi:10.1016/j.eswa.2025.126930 · Cited by 4
Robustness Evaluation of Knowledge Graph Embedding Models Under Non-targeted Attacks
Kapoor, S., Sharma, A., Röder, M., Demir, C., & Ngomo, A.-C. N. (2025) · Lecture notes in computer science · doi:10.1007/978-3-031-94575-5_15 · Cited by 4
An urban change detection method based on multimodal data and knowledge graph technology
Lu, K., Zhao, X., Li, M., Zhou, Y., & Zhang, B. (2025) · International Journal of Digital Earth · doi:10.1080/17538947.2025.2564902 · Cited by 4
“Additionally, TuckER, based on binary tensor Tucker Decomposition, is a linear representation model that achieves better results with fewer parameters.”
Boratto, L., Fenu, G., Marras, M., Medda, G., & Soccol, A. (2025) · ACM International Conference on Information and Knowledge Management (CIKM) · doi:10.1145/3746252.3761641 · Cited by 6
Offline and online coupled tensor factorization with knowledge graph
Lee, S.-J., Park, Y.-C., & Kang, U. (2025) · PLoS ONE · doi:10.1371/journal.pone.0336100 · Cited by 3
An explainable path reasoning framework for knowledge graph completion
Yu, H., Yu, H., Wang, W. W., et al. (2025) · Alexandria Engineering Journal · doi:10.1016/j.aej.2025.12.024 · Cited by 3
Huang, Q., Gao, S., Li, G., & Wang, W. (2026) · IEEE Transactions on Fuzzy Systems · doi:10.1109/tfuzz.2026.3655414 · Cited by 3
A Review of State-of-the-Art Deep Learning Models for Knowledge Graphs
Dube, B. M., & Fonou-Dombeu, J. V. (2026) · IEEE Access · doi:10.1109/access.2026.3660972 · Cited by 3
Generative Transformer with Knowledge-Guided Decoding for Academic Knowledge Graph Completion
Liu, X., Mao, S., Wang, X., & Bu, J. (2023) · Mathematics · doi:10.3390/math11051073 · Cited by 4
A lightweight CNN-based knowledge graph embedding model with channel attention for link prediction
Zhou, X., Guo, J., Jiang, L., Ning, B., & Wang, Y. (2023) · Mathematical Biosciences & Engineering · doi:10.3934/mbe.2023421 · Cited by 3
Entity and Entity Type Composition Representation Learning for Knowledge Graph Completion
Ni, R., Shibata, H., & Takama, Y. (2023) · Journal of Advanced Computational Intelligence and Intelligent Informatics · doi:10.20965/jaciii.2023.p1151 · Cited by 3
Hierarchical knowledge graph relationship prediction leverage of axiomatic fuzzy set graph structure
Fang, Y., Lang, Q., Lu, W., Liu, X., & Yang, J. (2024) · Expert Systems with Applications · doi:10.1016/j.eswa.2024.124090 · Cited by 3
XLORE 3: A Large-Scale Multilingual Knowledge Graph from Heterogeneous Wiki Knowledge Resources
Zeng, K., Jin, H., Lv, X., Zhu, F., Hou, L., Zhang, Y., Pang, F., Qi, Y., Liu, D., Li, J., et al. (2024) · ACM Transactions on Information Systems · doi:10.1145/3660521 · Cited by 3
“(2) Models based on tensor decomposition [7, 130].”
Designing a Knowledge Graph System for Digital Twin to Assess Urban Flood Risk
Feng, Y., Wang, Y., Xu, D., Zhang, X., & Jin, G. (2024) · Lecture notes in computer science · doi:10.1007/978-981-97-2303-4_13 · Cited by 3
Meta-collaboration-based semantic contrast for inductive knowledge representation learning
Fang, H., Cheng, K. K., Zhang, R., et al. (2024) · Expert Systems with Applications · doi:10.1016/j.eswa.2024.125421 · Cited by 3
Zero-Shot Relational Learning for Multimodal Knowledge Graphs
Cai, R., Pei, S., & Zhang, X. (2024) · IEEE International Conference on Big Data (BigData) · doi:10.1109/bigdata62323.2024.10825189 · Cited by 5
“Similar to TuckER [14], IMF [27] employs Tucker Decomposition on each modality to capture interactions between modalities.”
KG-CF: Knowledge Graph Completion with Context Filtering under the Guidance of Large Language Models
Zheng, Z., Dong, Y., Wang, S., Liu, H., Wang, Q., & Li, J. (2024) · IEEE International Conference on Big Data (BigData) · doi:10.1109/bigdata62323.2024.10826107 · Cited by 3
“Among them, RuleN [18] (rule-based) and GRAIL [19] (GNN-based) support both inductive and trans-ductive scenarios, while MINERVA [20] (reinforcement learning) and TuckER [21] (embedding-based) are limited to trans-ductive settings.”
Predicting Mobile App Usage With Context-Aware Dynamic Hypergraphs
Huang, Z., Li, T., Wang, X., Yang, K., Deng, C., Feng, J., & Li, Y. (2025) · IEEE Transactions on Mobile Computing · doi:10.1109/tmc.2025.3532992 · Cited by 4
“To fully exploit spatial features modeled in the knowledge graph, we employ TuckER [21], a state-of-the-art tensor factorization technique for knowledge graph embedding, mapping each entity to a low dimensional vector while preserving their semantic meanings.”
Reinforcement knowledge graph reasoning based on dual agents and attention mechanism
Yang, X.-H., Wang, T., Gan, J., et al. (2025) · Applied Intelligence · doi:10.1007/s10489-024-06162-x · Cited by 3
ZhangSun, J., Yu, Y., Zou, B., et al. (2025) · PeerJ Computer Science · doi:10.7717/peerj-cs.2522 · Cited by 3
“We further enhance this graph by integrating entity type embedding using TuckER ( Balažević, Allen & Hospedales, 2019 ).”
InfluentialModeling N-ary Relational Knowledge Bases with Tensor Decomposition
Liu, Y., Yao, Q., & Li, Y. (2025) · ACM Transactions on Intelligent Systems and Technology · doi:10.1145/3709002 · Cited by 3
“Based on the scoring function design, the typical works in binary relational KBs can be categorized into three groups: translational distance models [6, 19, 26, 50], neural network models [8, 37, 39], and tensor decomposition models [2, 22, 34, 43, 52].”
InfluentialJoshi, H. S., & Taherdoost, H. (2025) · International Conference on Multi-Agent Systems for Collaborative Intelligence (ICMSCI) · doi:10.1109/icmsci62561.2025.10894571 · Cited by 3
“This allows the embeddings to reflect varied semantic contexts based on the relation type [17].”
A detailed analysis into evaluation metrics for knowledge graph evaluation
Hassan, H., Elayidom, M. S., Irshad, M. R., & Chesneau, C. (2025) · International Journal of Data Science and Analytics · doi:10.1007/s41060-025-00871-3 · Cited by 3
AppGen: Mobility-Aware App Usage Behavior Generation for Mobile Users
Huang, Z., Li, T., & Li, Y. (2025) · IEEE Transactions on Mobile Computing · doi:10.1109/tmc.2025.3605939 · Cited by 3
“Then, we apply TuckER [34], a state-of-the-art tensor factorization technique, for knowledge graph embedding, which maps each entity to a low-dimensional vector while maintaining their semantic Authorized licensed use limited to the terms of the applicable license agreement with IEEE.”
Link Prediction in Knowledge Graphs with Numeric Triples Using Clustering
Bayrak, B., Choupani, R., & Doğdu, E. (2020) · IEEE International Conference on Big Data (BigData) · doi:10.1109/bigdata50022.2020.9378475 · Cited by 3
“For example, question answering [1], item recommendation [2], link prediction [3], searching [4], entity alignment [5], and more.”
BackgroundPath Ranking Model for Entity Prediction
Xiao, L., Yao, M., Zhuang, L., Li, H., & Wang, S. (2021) · IEEE International Conference on Multimedia and Expo (ICME) · doi:10.1109/icme51207.2021.9428138 · Cited by 3
“To demonstrate the effectiveness of the PRM in link prediction task, we select several involved state-of-the-art models for comparison, including two types of baselines: (1) KGE methods, which has been widely used in entity prediction tasks, such as: TransE [8], DistMult [9], ComplEx [14], RotatE [10], ConvE [15], R-G…”
MethodKnowledge Graph Completion Based on Multi-Relation Graph Attention Network
Wei, Q. (2022) · International Conference on Big Data and Education (ICBDE) · doi:10.1145/3524383.3524429 · Cited by 3
“[10] proposes a triplet embedding method based on tensor binary vector representation for link prediction, which obtained the better prediction results by projecting the triplet into another representation space; KGC based on GNN: Considering KGs as multi-relational directed graphs, scholars model KGs through GNN.”
MethodBastos, A., Singh, K., Nadgeri, A., Hoffart, J., Singh, M., & Suzumura, T. (2023) · International World Wide Web Conference (WWW) · doi:10.1145/3543507.3583308 · Cited by 3
“1), we use popular KG embedding methods from its various categories: (1) Translation: TransE [8], TransH [58], TransR [26] (2) Bilinear, Rotation, and Factorization: RotatE [44] TuckER [3], and ComplEx [48], (3) Neural Network based: ConvKB [32].”
MethodInfluentialIntent Classification by the Use of Automatically Generated Knowledge Graphs
Arčan, M., Manjunath, S. H., Robin, C., Verma, G., Pillai, D., Sarkar, S., Dutta, S., Assem, H., McCrae, J. P., & Paul Buitelaar, C. (2023) · Information · doi:10.3390/info14050288 · Cited by 3
Knowledge graph embedding by fusing multimodal content via cross-modal learning
Liu, S., Li, K., Wang, Y., Zhu, T., Li, J., & Chen, Z.-Y. (2023) · Mathematical Biosciences & Engineering · doi:10.3934/mbe.2023634 · Cited by 4
Knowledge Tensor-Aided Breast Ultrasound Image Assistant Inference Framework
Li, G., Xiao, L., Wang, G., Liu, Y., Liu, L., & Huang, Q. (2023) · Healthcare · doi:10.3390/healthcare11142014 · Cited by 3
“TuckER [27] decomposition is employed to reconstruct knowledge entities and relations to tensors.”
MethodFast and Slow Thinking: A Two-Step Schema-Aware Approach for Instance Completion in Knowledge Graphs
Yang, D., Qu, B., Rosso, P., & Cudré-Mauroux, P. (2023) · IEEE Transactions on Knowledge and Data Engineering · doi:10.1109/tkde.2023.3304137 · Cited by 4
“…works extend RESCAL by reducing the complexity of the models [26], [27] or of the training processes [28], by improving the model expressiveness [29], by capturing asymmetric relations [30], by considering unbalanced data in KGs [31], or by modeling non-linear relations using neural networks…”
BackgroundZhang, Q., Chen, B., & Liu, G. (2023) · Journal of Breath Research · doi:10.1088/1752-7163/acf065 · Cited by 3
“Then the above pathological relationship is sent to the pre-trained TuckER model to predict the tail entity and generate a supplement action space.”
InfluentialIntegrity verification for scientific papers: The first exploration of the text
Shi, X. D., Liu, Y., Liu, J., Cheng, Q., & Lu, W. P. (2023) · Expert Systems with Applications · doi:10.1016/j.eswa.2023.121488 · Cited by 3
Hierarchical Diachronic Embedding of Knowledge Graph Combined with Fragmentary Information Filtering
Liu, K., Wang, Z., Yang, Y., Huang, C., Niu, M., & Lü, Q. (2023) · Lecture notes in computer science · doi:10.1007/978-3-031-44216-2_36 · Cited by 3
Simple knowledge graph completion model based on PU learning and prompt learning
Duan, L., Wang, J., Luo, B., & Sun, Q. (2024) · Knowledge and Information Systems · doi:10.1007/s10115-023-02040-z · Cited by 3
Edge-enabled anomaly detection and information completion for social network knowledge graphs
Lu, F., Qin, H., & Qi, Q. (2024) · Wireless Networks · doi:10.1007/s11276-024-03671-y · Cited by 3
“Additionally, these methods incur significant computational and storage resource overheads, making deployment on edge devices impractical.”
Memory-Enhanced Knowledge Reasoning with Reinforcement Learning
Guo, J., Zhang, X., Liang, K., & Zhang, G. (2024) · Applied Sciences · doi:10.3390/app14073133 · Cited by 3
ShallowBKGC: a BERT-enhanced shallow neural network model for knowledge graph completion
Jia, N., & Yao, C. (2024) · PeerJ Computer Science · doi:10.7717/peerj-cs.2058 · Cited by 4
Incorporating structural knowledge into language models for open knowledge graph completion
Song, X., Wang, Y., Zhou, B., Wang, H., Huang, Y., & Gao, L. (2024) · World Wide Web · doi:10.1007/s11280-024-01314-y · Cited by 3
“It contains: (1) KG Embedding-based methods, we evaluate the TransE [3], RotatE [4], TuckER [33], ConvE [36] and HJE [37], where the HJE is the latest baseline.”
Dang, H., & Nguyen, T. T. M. (2025) · Knowledge and Information Systems · doi:10.1007/s10115-025-02364-y · Cited by 3
“Tucker [44] uses the Tucker decomposition technique to separate a set of vector embeddings representing entities and relations.”
What Comes After Transformers? A Selective Survey Connecting Ideas in Deep LearningGPT
Schneider, J. (2025) · Lecture notes in computer science · doi:10.1007/978-3-031-87330-0_4 · Cited by 14
RankT: Ranking-Triplets-based adversarial learning for knowledge graph link prediction
Zhu, J., Zhang, X., & Ding, X. D. (2025) · Data & Knowledge Engineering · doi:10.1016/j.datak.2025.102463 · Cited by 4
Knowledge Base Completion with transfer learning using BERT and fastText
Thi, T.-A. N., Vuong, T.-H., Le, T.-H., Phan, X.-H., Le, T.-T., & Ha, Q.-T. (2022) · International Conference on Knowledge and Systems Engineering/Proceedings · doi:10.1109/kse56063.2022.9953802 · Cited by 3
“Missing facts can also be inferred from existing ones using knowledge base completion algorithms, such as ConvE [6], TuckER [7], 5 (cid:1) E [8], and the transfer learning-based model by Kocijan and Lukasiewicz [2].”
BackgroundUnsupervised Type Constraint Inference in Bilinear Knowledge Graph Completion Models
Lu, Y., & Ichise, R. (2021) · IEEE International Conference on Big Knowledge (ICBK) · doi:10.1109/ickg52313.2021.00012 · Cited by 3
A Two-Stage Framework for Directed Hypergraph Link Prediction
Xiao, G., Liao, J., Tan, Z., et al. (2022) · Mathematics · doi:10.3390/math10142372 · Cited by 3
“For predicting the existence of hyperlinks, we exploit Tucker decomposition to model hyperlinks, which, to the best of our knowledge, has not been applied to hypergraphs except simple graphs [15].”
BackgroundMethodInfluentialRule Induction in Knowledge Graphs Using Linear Programming
Dash, S., & Goncalves, J. (2023) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v37i4.25541 · Cited by 5
“We ran ComplEx-N3 and TuckER on our machines with the best published hyperparameters (if available).”
BackgroundMethodInfluentialGK: A Graphormer Architecture Based Knowledge Graph Reasoning Model for Link Prediction
Wu, Y., & Zhou, J. (2022) · IEEE International Conference on Parallel & Distributed Processing with Applications, Big Data & Cloud Computing, Sustainable Computing & Communications, Social Computing & Networking (ISPA/BDCloud/SocialCom/SustainCom) · doi:10.1109/ispa-bdcloud-socialcom-sustaincom57177.2022.00019 · Cited by 2
AttFGCN: A GCN-Based Method Using Attention Flow for Knowledge Graph Completion
Yu, M., Gao, F., Zhao, M., et al. (2024) · Lecture notes in computer science · doi:10.1007/978-981-97-5562-2_27 · Cited by 2
A Comprehensive Survey of Few-shot Information Networks
Zheng, X., Che, F., & Tao, J. (2025) · Machine Intelligence Research · doi:10.1007/s11633-023-1470-4 · Cited by 2
Subgraph retrieval and link scoring model for multi-hop question answering in knowledge graphs
Zhou, C., Ying, W., Zhong, S., Gong, S., & Yan, H. (2025) · Applied Intelligence · doi:10.1007/s10489-024-05935-8 · Cited by 2
“Performing tensor analysis on the above third-order tensor is equivalent to mining the implicit relationships between all entities on the KGs. Inspired by Balazevic et al. [27], we constructed a KGs link prediction model that cleverly extends to downstream tasks.”
Vu, T., & Le, T. (2025) · Knowledge-Based Systems · doi:10.1016/j.knosys.2025.113321 · Cited by 2
PR-KGC: Text-enhanced Knowledge Graph Completion with Pair-wise Re-ranking
Gao, M., Zhang, F., Chen, W., Wang, T., Zheng, J., & Yao, J. (2025) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp49660.2025.10889671 · Cited by 3
“(1) Triple-based methods: TransE [4], ConvE [23], WWV [24], TuckER [8] and RotatE [7].”
Wang, T., & Shen, B. (2025) · Complex & Intelligent Systems · doi:10.1007/s40747-025-01860-6 · Cited by 2
“TuckER [21]: a simple but powerful linear model based on the decomposition of binary tensor representations.”
Two-layer tensor decomposition for temporal kge graph completion
Yue, L., Ren, Y., Zeng, Y., et al. (2025) · Expert Systems with Applications · doi:10.1016/j.eswa.2025.127953 · Cited by 2
Multi-source information diffusion model for conversation recommender system
Liu, F., Huang, X., & Liu, H. (2025) · Evolving Systems · doi:10.1007/s12530-025-09692-y · Cited by 3
“The cross-associated fusion mechanism is an extension of Tucker decomposition (Balazevic et al. 2019).”
ReaLitE: Enrichment of Relation Embeddings in Knowledge Graphs Using Numeric Literals
Klironomos, A., Zhou, B., Zheng, Z., Mohamed, G.-E., Paulheim, H., & Kharlamov, E. (2025) · Lecture notes in computer science · doi:10.1007/978-3-031-94575-5_3 · Cited by 2
“Similar to literal-aware KGE model [31], we combine ReaLitE with five models: TransE [7], DistMult [34], ComplEx [29], RotatE [27], and TuckER [3].”
Knowledge graph completion based on iteratively learning embeddings and noise-aware rules
Zhang, J., & Shen, B. (2025) · Journal of King Saud University - Computer and Information Sciences · doi:10.1007/s44443-025-00148-6 · Cited by 2
“Tucker (Balaževi´c et al. 2019) factorizes a tensor into a core tensor multiplied by matrices along each mode, allowing it to better capture interactions between entities and relations.”
Zhao, L., Wang, B., Gao, J., Li, X., Hu, Y., & Yin, B. (2025) · IEEE Transactions on Big Data · doi:10.1109/tbdata.2025.3600014 · Cited by 2
Identification of Missing Knowledge in MBSE System Models Using Graph‐Based Machine Learning
Karagoz, E., Fischer, O. J. P., & Mavris, D. N. (2025) · Systems Engineering · doi:10.1002/sys.70013 · Cited by 3
“Some prominent tensor decomposition models for link prediction include DistMult [52], CompIEx [53], Analogy [54], SimpIE [55], HoIE [56], and Tucker Decomposition [57].”
Ren, W., Zhao, L. H., & Li, W. (2026) · Knowledge and Information Systems · doi:10.1007/s10115-025-02669-y · Cited by 2
“TuckER [23] extends the semantic matching model to the three-dimensional tensor decomposition of knowledge graph triple binary tensor representation.”
Liu, Y., Wang, Q., Chen, W., & Shen, T. (2026) · Complex & Intelligent Systems · doi:10.1007/s40747-025-02225-9 · Cited by 3
“TuckER [31], based on the tensor decomposition framework, unifies the modeling of multimodal interactions through a core tensor.”
Securely Answering Multi-Hop Questions on the Joint of Private and Public Knowledge Graphs
Li, S., An, B., Huo, X., Ma, J., Zhang, Y., Cai, L., Wang, P., & Yan, C.-B. (2026) · IEEE Transactions on Automation Science and Engineering · doi:10.1109/tase.2025.3649996 · Cited by 2
“For example, there are the translation-based model [23], [24], [25], [26] represented by TransE [16] and the tensor decomposition model [27], [28], [29] represented by RESCAL [30].”
FLS: fusion space and smooth lie group-driven temporal knowledge graph embedding
Chen, Z., Wu, J., & Wang, K. (2026) · International Journal of Machine Learning and Cybernetics · doi:10.1007/s13042-026-03075-8 · Cited by 2
Transductive Data Augmentation with Relational Path Rule Mining for Knowledge Graph Embedding
Hirose, Y., Shimbo, M., & Watanabe, T. (2021) · IEEE International Conference on Big Knowledge (ICBK) · doi:10.1109/ickg52313.2021.00057 · Cited by 2
“RESCAL [3] is the first bilinear KG embedding model and TuckER [5] is a recently proposed model shown to perform better.”
MethodInfluentialKnowledge Graph Embedding with Direct and Disentangled Neighborhood Representation Attention Network
Yu, R., Gao, S., Yu, J. X., et al. (2022) · Lecture notes in computer science · doi:10.1007/978-3-031-10983-6_22 · Cited by 2
QuatCNNEx: A Knowledge Graph Embedding Model based on Quaternion Convolutional Neural Network
Su, H., Xiong, W., Wu, Y., Chen, L., & Zhong, Z. (2022) · International Conference on Big Data, Artificial Intelligence and Internet of Things Engineering (ICBAIE) · doi:10.1109/icbaie56435.2022.9985846 · Cited by 2
“In the experiment, we compared QuatCNNEx to various baselines, including translation distance models such as TransE[1], RotatE[8], and semantic matching models: DistMult[3], ComplEx[6], ConvE[13], QuatE[4], TuckER[12], ConEx[5].”
MethodComplex Knowledge Graph Embeddings Based on Convolution and Translation
Shi, L., Zhao, Y., Ji, Z., & Ганчев, И. (2023) · Mathematics · doi:10.3390/math11122627 · Cited by 3
Metadata
- Authors
- Balazevic, I., Allen, C., & Hospedales, T.
- Affiliations
- University of Edinburgh (GB)
- Venue
- Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP)
- Published
- 2019
- Pages
- 5184-5193
- OpenAlex
- W2914592219
- Cited by
- 696
- References
- 0
- Citations by year
- 2026: 81 · 2025: 125 · 2024: 165 · 2023: 137 · 2022: 109 · 2021: 55 · 2020: 21 · 2019: 3
- FWCI
- 17.97
- Type
- conference paper
- Language
- en
- Open access
- gold
- doi.org
- Licence
- cc-by
- Locations
- arXiv (Cornell University) (submittedVersion)
- Edinburgh Research Explorer (University of Edinburgh) (submittedVersion, cc-by)
- arXiv (Cornell University)
- Funders
- University of Edinburgh
- Engineering and Physical Sciences Research Council (EP/L016427/1)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Advanced Graph Neural Networks
- Physical Sciences › Computer Science › Artificial Intelligence › Topic Modeling
- Physical Sciences › Mathematics › Computational Mathematics › Tensor decomposition and applications
- Keywords
- knowledge graph completion, tensor decomposition, Tucker, knowledge graphs
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
A language model drew the claims from the paper on 2026-10-23, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-23 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (gpt-6-sol).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
