Observed versus latent features for knowledge base and text inference
Open accessAbstract
In this paper we show the surprising effectiveness of a simple observed features model in comparison to latent feature models on two benchmark knowledge base completion datasets, FB15K and WN18.We also compare latent and observed feature models on a more challenging dataset derived from FB15K, and additionally coupled with textual mentions from a web-scale corpus.We show that the observed features model is most effective at capturing the information present for entity pairs with textual relations, and a combination of the two combines the strengths of both model types.
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
On FB15k-237, the key of link prediction is to model and infer the symmetry/antisymmetry and composition patterns.
“FB15k-237 (Toutanova & Chen, 2015) is a subset of FB15k, where inverse relations are deleted. Therefore, the key of link prediction on FB15k-237 boils down to model and infer the symmetry/antisymmetry and composition patterns.”
From RotatE: Knowledge Graph Embedding by Relational Rotation in Complex Space (2019) · cites this paper as Toutanova & Chen, 2015
Section 4.3 describes FB15KSelected as a subset of FB15K that retains one relation from each detected inverse or near-duplicate set, leaving 237 relations. Section 4.2 says the resulting task requires non-trivial inference after trivial cases involving symmetry or inverse relations are addressed. Neither section says that link prediction on this dataset boils down to modeling symmetry, antisymmetry, and composition patterns; composition and antisymmetry are not established as the key patterns. Page numbers are unavailable in our copy. (LLM, conf=0.86)
Experimental — accuracy not yet measuredOn FB15k, the key of link prediction is to model and infer the symmetry/antisymmetry and inversion patterns.
“Toutanova & Chen (2015) showed that almost 81% of the test triplets ( x , r , y ) can be inferred via a directly linked triplet ( x , r ′ , y ) or ( y , r ′ , x ) . Therefore, the key of link prediction on FB15k is to model and infer the symmetry/antisymmetry and inversion patterns.”
From RotatE: Knowledge Graph Embedding by Relational Rotation in Complex Space (2019) · cites this paper as Toutanova & Chen (2015)
Section 4.2 reports that almost 81% of FB15K test triples have an entity pair directly linked in the training graph, and connects easy inferences to relation symmetry or inverse relations. The paper does not identify antisymmetry as a pattern behind that figure, nor establish that modeling these patterns is the key to FB15K link prediction overall. (LLM, conf=0.83)
Experimental — accuracy not yet measuredWN18RR was built to be a more challenging dataset than WN18 by applying a pipeline similar to the one employed for FB15k-237.
“To resolve that, they build the far more challenging WN18RR dataset by applying a pipeline similar to the one employed for FB15k-237 [57].”
From Knowledge Graph Embedding for Link Prediction (2021) · cites this paper as [57]
Section 4.3 describes a pipeline that removes duplicate and inverse relations to create the more challenging FB15KSelected dataset, with 237 relations. Section 4.2 discusses WN18, but the paper does not say that WN18RR was built using a similar pipeline or assess how challenging WN18RR is relative to WN18. (LLM, conf=0.88)
Experimental — accuracy not yet measuredFB15k suffers from test leakage, consisting in test data being seen by models at training time, due to relations that are near-identical or the inverse of one another.
“FB15k-237 is a subset of FB15k built by Toutanova and Chen [57], inspired by the observation that FB15k suffers from test leakage , consisting in test data being seen by models at training time. In FB15k this issue is due to the presence of relations that are near-identical or the inverse of one another.”
From Knowledge Graph Embedding for Link Prediction (2021) · cites this paper as [57]
FB15k-237 was built to be a more challenging dataset by selecting facts from FB15k involving the 401 largest relations and removing all equivalent or inverse relations.
“FB15k-237 was built to be a more challenging dataset: the authors first selected facts from FB15k involving the 401 largest relations and removed all equivalent or inverse relations.”
From Knowledge Graph Embedding for Link Prediction (2021) · cites this paper as [57]
A simple model based on observable features can easily reach state-of-the-art performance on FB15k.
“In order to assess the severity of this problem, Toutanova and Chen have shown that a simple model based on observable features can easily reach state-of-the-art performance on FB15k.”
From Knowledge Graph Embedding for Link Prediction (2021) · cites this paper as [57]
In FB15k-237, none of the entities connected in the training set are also directly linked in the validation and test sets.
“In order to filter away all trivial triples, they also ensured that none of the entities connected in the training set are also directly linked in the validation and test sets.”
From Knowledge Graph Embedding for Link Prediction (2021) · cites this paper as [57]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Cites
Incorporating Vector Space Similarity in Random Walk Inference over Knowledge Bases
Gardner, M., Talukdar, P., Krishnamurthy, J., & Mitchell, T. (2014) · Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing (EMNLP) · doi:10.3115/v1/d14-1044 · Cited by 189 · see citing works
Flagged first, then by citations
Research works printed but not linked: 18
- Sören Auer, Christian Bizer, Georgi Kobilarov, Jens Lehmann, Richard Cyganiak, and Zachary Ives. 2007. Dbpedia: A nucleus for a web of open data. Springer.
- Kurt Bollacker, Colin Evans, Praveen Paritosh, Tim Sturge, and Jamie Taylor. 2008. Freebase: a col- laboratively created graph database for structuring human knowledge. In Proceedings of the 2008 ACM SIGMOD international conference on Management of data, pages 1247-1250. ACM.
- Antoine Bordes, Jason Weston, Ronan Collobert, and Yoshua Bengio. 2011. Learning structured embed- dings of knowledge bases. In Conference on Artifi- cial Intelligence, number EPFL-CONF-192344.
- 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 (NIPS).
- Kai-Wei Chang, Wen-tau Yih, Bishan Yang, and Christopher Meek. 2014. Typed tensor decompo- sition of knowledge bases for relation extraction. In Empirical Methods in Natural Language Processing (EMNLP).
- Xin Dong, Evgeniy Gabrilovich, Geremy Heitz, Wilko Horn, Ni Lao, Kevin Murphy, Thomas Strohmann, Shaohua Sun, and Wei Zhang. 2014. Knowledge vault: A web-scale approach to probabilistic knowl- edge fusion. In International Conference on Knowl- edge Discovery and Data Mining (KDD).
- Matt Gardner, Partha Pratim Talukdar, Bryan Kisiel, and Tom Mitchell. 2013. Improving learning and inference in a large knowledge-base using latent syntactic cues. In Empirical Methods in Natural Language Processing (EMNLP).
- Ni Lao, Tom Mitchell, and William W Cohen. 2011. Random walk inference and learning in a large scale knowledge base. In Empirical Methods in Natural Language Processing (EMNLP).
- Ni Lao, Amarnag Subramanya, Fernando Pereira, and William W Cohen. 2012. Reading the web with learned syntactic-semantic inference rules. In Empirical Methods in Natural Language Process- ing and Computational Natural Language Learning (EMNLP/CoNLL).
- Dong C Liu and Jorge Nocedal. 1989. On the limited memory bfgs method for large scale optimization. Mathematical programming, 45(1-3):503-528.
- Maximilian Nickel, Volker Tresp, and Hans-Peter Kriegel. 2011. A three-way model for collective learning on multi-relational data. In Proceedings of the 28th international conference on machine learn- ing (ICML-11), pages 809-816.
- In Z. Ghahramani, M. Welling, C. Cortes, N.D. Lawrence, and K.Q. Weinberger, editors, Advances in Neural Information Processing Systems 27, pages 1179-1187.
- Maximilian Nickel, Kevin Murphy, Volker Tresp, and Evgeniy Gabrilovich. 2015. A review of re- lational machine learning for knowledge graphs: From multi-relational link prediction to automated knowledge graph construction. arXiv preprint arXiv:1503.00759. · arXiv
- Sebastian Riedel, Limin Yao, Benjamin M. Marlin, and Andrew McCallum. 2013. Relation extraction with matrix factorization and universal schemas. In North American Chapter of the Association for Computational Linguistics: Human Language Tech- nologies (NAACL-HLT).
- Martin Riedmiller and Heinrich Braun. 1993. A direct adaptive method for faster backpropagation learn- ing: The rprop algorithm. In Neural Networks, 1993., IEEE International Conference on, pages 586-591. IEEE.
- Fabian M Suchanek, Gjergji Kasneci, and Gerhard Weikum. 2007. Yago: a core of semantic knowl- edge. In Proceedings of the 16th international con- ference on World Wide Web, pages 697-706. ACM.
- Zhen Wang, Jianwen Zhang, Jianlin Feng, and Zheng Chen. 2014. Knowledge graph embedding by trans- lating on hyperplanes. In Proceedings of the Twenty- Eighth AAAI Conference on Artificial Intelligence, pages 1112-1119.
- 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 (ICLR).
Cited by
Sorted by citations
Modeling Relational Data with Graph Convolutional Networks
Schlichtkrull, M., Kipf, T. N., Bloem, P., van den Berg, R., Titov, I., & Welling, M. (2018) · Lecture Notes in Computer Science · doi:10.1007/978-3-319-93417-4_38 · Cited by 5,570 · see citing works
A Survey on Knowledge Graphs: Representation, Acquisition, and Applications
Ji, S., Pan, S., Cambria, E., Marttinen, P., & Yu, P. S. (2022) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2021.3070843 · Cited by 2,947 · see citing works
RotatE: Knowledge Graph Embedding by Relational Rotation in Complex Space
Sun, Z., Deng, Z.-H., Nie, J.-Y., & Tang, J. (2019) · arXiv · doi:10.48550/arxiv.1902.10197 · Cited by 1,414 · see citing works
A Novel Embedding Model for Knowledge Base Completion Based on Convolutional Neural Network
Nguyen, D. Q., Nguyen, T. D., Nguyen, D. Q., & Phung, D. (2018) · Proceedings of the 2018 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 2 (Short Papers) · doi:10.18653/v1/n18-2053 · Cited by 923 · see citing works
Knowledge Graph Embedding for Link Prediction
Rossi, A., Barbosa, D., Firmani, D., Matinata, A., & Merialdo, P. (2021) · ACM Transactions on Knowledge Discovery from Data · doi:10.1145/3424672 · Cited by 601 · see citing works
“is also applies to FB15k-237, that was obtained by just sampling FB15k further [57].”
BackgroundMethodInfluentialLow-Dimensional Hyperbolic Knowledge Graph Embeddings
Chami, I., Wolf, A., Juan, D.-C., Sala, F., Ravi, S., & Ré, C. (2020) · Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics · doi:10.18653/v1/2020.acl-main.617 · Cited by 384 · see citing works
Query2box: Reasoning over Knowledge Graphs in Vector Space using Box Embeddings
Ren, H., Hu, W., & Leskovec, J. (2020) · arXiv · doi:10.48550/arxiv.2002.05969 · Cited by 135 · see citing works
Pretrain-KGE: Learning Knowledge Representation from Pretrained Language Models
Zhang, Z., Liu, X., Zhang, Y., Su, Q., Sun, X., & He, B. (2020) · Findings of the Association for Computational Linguistics: EMNLP 2020 · doi:10.18653/v1/2020.findings-emnlp.25 · Cited by 52 · see citing works
Modeling Heterogeneous Hierarchies with Relation-specific Hyperbolic Cones
Bai, Y., Ying, R., Ren, H., & Leskovec, J. (2021) · arXiv · doi:10.48550/arxiv.2110.14923 · Cited by 31 · see citing works
“We use four knowledge graph benchmarks (Table 1): WordNet lexical knowledge graph (WN18RR [5, 17]), drug knowledge graph (DDB14 [18]), and a KG capturing common knowledge (FB15k-237 [31]).”
MethodBenchmark Data Contamination of Large Language Models: A Survey
Xu, C., Guan, S., Greene, D., & Kechadi, M.-T. (2024) · arXiv · doi:10.48550/arxiv.2406.04244 · Cited by 22 · see citing works
Symmetry-Aware Knowledge Graph Compression for LLM Grounding via Symbolic Retrieval
Prakash, V., & Dontabhaktuni, J. (2026) · doi:10.2139/ssrn.6826298
- Authors
- Toutanova, K., & Chen, D.
- Affiliations
- Microsoft (United States) (US)
- Venue
- Proceedings of the 3rd Workshop on Continuous Vector Space Models and their Compositionality
- Published
- 2015
- Pages
- 57-66
- OpenAlex
- W2250184916
- Cited by
- 1,070
- References
- 1
- Citations by year
- 2026: 69 · 2025: 127 · 2024: 171 · 2023: 162 · 2022: 148 · 2021: 136 · 2020: 118 · 2019: 70 · 2018: 37 · 2017: 23 · 2016: 6 · 2015: 3
- FWCI
- 12.96
- Type
- conference paper
- Language
- en
- Open access
- gold
- doi.org
- Licence
- cc-by
- Funders
- Microsoft Research
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Advanced Graph Neural Networks
- Physical Sciences › Computer Science › Artificial Intelligence › Topic Modeling
- Physical Sciences › Computer Science › Artificial Intelligence › Semantic Web and Ontologies
- Keywords
- knowledge base completion, latent features, textual relations, entity pairs, relation prediction
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.
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
