Pith. sign in

REVIEW 4 major objections 5 minor 35 references

A Semantic Partitioning Method for Large-Scale Training of Knowledge Graph Embeddings

T0 review · 4 major / 5 minor · reviewed 2026-08-10 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read This paper claims that partitioning knowledge-graph triples by the class of the head entity lets embeddings train in parallel without sacrificing link prediction quality.

desk verdict Simple ontology-based partitioning idea, honestly reported, but the semantic claim is untested and the link prediction evidence is mostly noise. read the letter →

arxiv 2501.04613 v1 pith:SIGR6QUC submitted 2025-01-08 cs.LG cs.DC

Yuhe Bai This is my paper · ORCID
classification cs.LGcs.DC
keywords knowledgegraphembeddingssemanticpartitioningontology-awareparalleltraininglinkpredictionentitytypingFreebasedistributedembedding
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper claims that when a knowledge graph is too large to train embeddings on in a single pass, the right way to cut it into pieces is by the semantic class of the head entity rather than at random. This class-based partitioning keeps the scoring functions of existing embedding models untouched, so it can be dropped into most current methods, and it turns the class information of the ontology into an organizing principle for parallel training. On the Freebase-derived benchmarks FB15K and FB15K-237, using TransE, DistMult, and ComplEx, the preliminary results are generally comparable to random partitioning, with DistMult showing the clearest improvement. The paper presents this as a first step toward embedding training that is both scalable and semantically informative, with entity typing as the planned test of whether the extra semantic information actually helps.

What carries the argument

The machinery is ontology-based semantic partitioning. A knowledge graph is represented as $KG=(E,R,T)$ with entities, relations, and fact triplets; the ontology supplies class labels and rdf:type statements. Each entity is assigned to its lowest-level class to maximize the number of partitions, and every fact triplet is routed to the partition named by its head entity's class. This cut is what lets the method keep any existing model's scoring function while injecting class-level semantics into training, and it is what makes the workload parallelizable. The evaluation runs the partitioned training through a parallel knowledge-graph embedding library using TransE, DistMult, and ComplEx with the library's best hyperparameters on 64 CPUs.

What would settle it

Run the same parallel training pipeline on the same datasets with semantically partitioned and randomly partitioned triplets, while also reporting how overlapping entity embeddings are merged; if random partitioning matches or beats semantic partitioning on both link prediction metrics and entity-typing F1 across class levels, then the class information in the cut is not what drives the results.

Watch

Extended reading notes

Core claim

The paper's central claim is that ontology information should participate in large-scale knowledge graph embedding through the partition step, not through a new scoring function. Concretely, each entity is assigned to its most specific (lowest) class, and each fact triplet $\langle h,r,t\rangle$ is placed in the partition of its head entity's class, so every partition contains triples that share a semantic category. Training these partitions in parallel with off-the-shelf models then yields embeddings that are meant to carry more semantic information while matching the link prediction quality of a centralized or randomly partitioned run. The reported numbers bear this out unevenly: on FB15K-237, DistMult improves from MRR 25.4 to 26.1 and Hits@10 from 43.7 to 44.4, TransE is nearly unchanged, and ComplEx is slightly worse; on FB15K, DistMult again improves on both metrics, with TransE and ComplEx close to random partitioning. The paper states these are preliminary results and that the benefit is model-dependent.

Load-bearing premise

The load-bearing premise is that training on class-based partitions produces valid embeddings even though the same entity appears in many partitions, and the paper does not specify how the shared entity vectors are kept consistent across partitions.

Editorial extensions

If this is right

  • Very large knowledge graphs that do not fit on one machine could be trained in parallel by class-based partitions while keeping the scoring function of the chosen embedding model unchanged.
  • The method is a preprocessing layer: any translational, tensor-decomposition, or deep knowledge graph embedding model should be able to adopt it without retraining internals.
  • Partitioning by the most specific class increases the number of partitions and the parallelism, while concentrating triples with shared semantic categories in each partition.
  • If the semantic-information claim holds, downstream tasks such as entity typing and entity alignment should benefit even on benchmarks where link prediction is flat.
  • The best partition-and-negative-sampling configuration is likely model- and dataset-dependent, matching the paper's observation that DistMult gains while ComplEx loses.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • A broader design principle is suggested: the partition boundary is itself an inductive bias, so the same idea could be tested with relation-based cuts, hierarchy-aware cuts, or cuts that respect subclass edges, not just head-entity class.
  • A decisive test the paper announces but does not run is entity typing F1 at several class levels; if low-level classes improve there while link prediction is flat, the semantic-information claim would be supported independently of ranking metrics.
  • The unresolved mechanism is how embeddings of entities shared across partitions stay consistent; until that synchronization is specified, some of the observed behavior may come from the parallel training infrastructure rather than from the semantic content of the cut.
  • The method could combine with ontology-embedding techniques that encode subclass structure, using the class partition to scale those techniques to graphs much larger than the centralized datasets where they are usually evaluated.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. The paper proposes a semantic partitioning method for large-scale knowledge graph embedding training. Entities are assigned to their lowest ontology class, fact triplets are partitioned according to the class of the head entity, and the resulting partitions are trained in parallel with DGL-KE using TransE, DistMult, and ComplEx. The authors compare their method with random partitioning on FB15K and FB15K-237, reporting MRR and Hits@10 in Table 2. The abstract claims that the method incorporates ontology information, injects semantic information into embeddings, and performs well on popular benchmarks, but the paper also states that the current results are preliminary and that entity classification, which would test the semantic claim, is left to future work.

Significance. If the central claims were convincingly supported, the method would be an attractive low-cost preprocessing step: it is model-agnostic, uses existing ontology data, and fits naturally into parallel training. The paper's main strength is the simplicity of the proposed partitioning rule and the use of a standard, publicly available training library (DGL-KE). However, the evidence presented is not sufficient to establish the claimed benefits. The abstract's statement that the algorithm 'performs well on several popular benchmarks' is not supported by the mixed results in Table 2, and the paper's core assertion that the embeddings contain 'more semantic information' is never directly tested. The current contribution is therefore best viewed as an early-stage idea requiring substantial additional evaluation.

major comments (4)
  1. [Section 6, Table 2] The central claim that the method 'performs well on several popular benchmarks' is not supported by the reported numbers. On FB15K-237, semantic partitioning improves over random partitioning only for DistMult (MRR 26.1 vs 25.4), while ComplEx is slightly worse (25.3 vs 25.4) and TransE gains only 0.1 MRR. On FB15K, TransE results are identical to random partitioning (69.2 MRR and 85.9 Hits@10), and ComplEx is worse by 1.0 MRR and 0.5 Hits@10. The only consistent positive result is DistMult, and no error bars, seeds, or repeated runs are reported. The paper's own Section 6 text acknowledges that the method is model-dependent and inferior to random partitioning for ComplEx, which contradicts the abstract's blanket claim of good performance.
  2. [Section 5, Section 6, Section 7] The claimed mechanism, that class-based partitioning yields embeddings with more semantic information, is never directly evaluated. Section 5 states that 'the obtained embeddings are then utilized for entity classification tasks,' but no precision, recall, or F1 results appear anywhere in the paper; Sections 6 and 7 list entity classification as future work. Without any entity classification or other semantic evaluation, the assertion that the method 'preserves more semantic information' is an unsupported conjecture rather than a demonstrated result.
  3. [Section 5] The paper does not explain how embeddings of entities appearing in multiple partitions are updated, synchronized, or merged during parallel training. Section 5 says only that each fact triplet is classified based on the class of its head entity, but an entity can belong to several classes and appear in several partitions. Since the parallel implementation (DGL-KE) has its own consistency mechanisms, the observed differences in Table 2 cannot be attributed to semantic grouping without a description of how entity embeddings are reconciled across partitions and without verifying that the final vectors are consistent. The class-frequency imbalance shown in Figure 1 also means semantic and random partitions have very different sizes, so partition-size effects are a plausible confound.
  4. [Section 2 and Section 4] The paper lists subgraph selection as the first sub-problem and as a proposed contribution, but no experiment or evaluation of subgraph selection is provided. Section 2 defines sub-problem (1) as finding the subgraph of size p*s that maximizes prediction quality, and Section 4 claims a subgraph selection method, yet all reported results concern full-graph partitioning for link prediction. This is an untested part of the stated contribution and should be either evaluated or explicitly removed from the paper's claims.
minor comments (5)
  1. [Table 2] The citations for the 'First Proposed' rows are incorrect: DistMult is cited as [4] and ComplEx is also cited as [4], but reference [4] is the ConvE paper. DistMult should cite Yang et al. [12] and ComplEx should cite Trouillon et al. [27].
  2. [Section 6] The 'First Proposed' values in Table 2 are taken from prior studies and may have been obtained with different training configurations than the DGL-KE-based random and semantic partitioning runs. A fair comparison should either use the same backbone and hyperparameters for all rows or clearly state that the first-proposed rows are literature reference points, not controlled comparisons.
  3. [Figure 1] Figure 1, 'Freebase class frequency analysis,' has no visible axis labels or class names, and the text does not describe how class frequency was computed. For a reader to judge the claimed class imbalance, the figure should include labeled axes and a short description of the counting procedure.
  4. [Section 5] The last paragraph of Section 5 says that the proposed partitioning is 'a preliminary, simple semantic partitioning method and requires further refinement.' This caveat is important, but the paper does not quantify how much refinement is needed or which aspects of the method are tentative.
  5. [Section 6] The sentence 'If we look at the preliminary link prediction results we obtained now, our method performs well in overall' is difficult to reconcile with the immediately preceding sentences reporting model-dependent and partly inferior results. Please either rephrase the conclusion to reflect the mixed evidence or provide statistical support for the claim.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the partitioning heuristic is fixed and evaluated on external link-prediction benchmarks.

full rationale

The paper's derivation chain is a fixed preprocessing heuristic: assign each entity to its lowest class, partition each fact triplet by its head entity's class, train standard KGE models (TransE, DistMult, ComplEx) with DGL-KE, and compare against random partitioning on the standard FB15K and FB15K-237 test splits. No parameter is fitted to the evaluation data, no test labels are used to choose the partition rule, and no equation in the paper reduces the reported MRR or Hits@10 values to the method's own inputs. The claim that class-based partitioning 'includes more semantic information' is asserted rather than measured, and the paper itself notes that results are model-dependent; but an unsupported or weakly supported claim is an evidence problem, not a circularity problem. There are no self-citations carrying load, no imported uniqueness theorems, and no known result being renamed. Accordingly, no circular step can be exhibited with the required specificity, and the circularity score is 0.

Assumptions & free parameters 1 free parameters · 3 assumptions · 0 invented entities

The method introduces no new entities, forces, or conserved quantities. Its load-bearing inputs are class labels from an external ontology and a heuristic assignment rule; the ledger reflects those two dependencies plus the untested coherence assumption.

free parameters (1)
  • Entity-to-class assignment rule (lowest class)
    Section 5 states 'we assign each entity to its lowest class, which increases the number of partitions and allows for higher parallelism.' This is a hand-chosen design choice, not a learned or fitted value, and the paper provides no evidence that lowest-class assignment is optimal.
assumptions (3)
  • domain assumption Benchmark entities have usable Freebase class information
    Section 5 says they used 'facts, classes, and rdf:type information from the Freebase knowledge base,' but the paper does not describe how FB15K and FB15K-237 entities are mapped to Freebase types or how missing or multiple types are handled.
  • domain assumption Partitioning by head entity class preserves embedding quality without special synchronization
    Section 5 assigns each fact triplet to the class of its head entity, but an entity can appear in many partitions. The paper does not specify how overlapping entity embeddings are updated or merged during parallel training, yet assumes the resulting embeddings are valid.
  • ad hoc to paper Lower-level classes produce more semantically coherent partitions
    Section 5: 'we also believe that this should enhance the embedding quality because each partition contains more specific semantic information.' This is stated as a belief, and the paper does not test it directly.

how reviews work

0 comments
Cite this review

Pith. "Pith review of A Semantic Partitioning Method for Large-Scale Training of Knowledge Graph Embeddings." pith.science (2026). https://pith.science/paper/SIGR6QUC

@misc{pith2026250104613,
  author       = {Pith},
  title        = {Pith review of: A Semantic Partitioning Method for Large-Scale Training of Knowledge Graph Embeddings},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/SIGR6QUC}},
  note         = {Machine review of arXiv:2501.04613}
}
read the original abstract

In recent years, knowledge graph embeddings have achieved great success. Many methods have been proposed and achieved state-of-the-art results in various tasks. However, most of the current methods present one or more of the following problems: (i) They only consider fact triplets, while ignoring the ontology information of knowledge graphs. (ii) The obtained embeddings do not contain much semantic information. Therefore, using these embeddings for semantic tasks is problematic. (iii) They do not enable large-scale training. In this paper, we propose a new algorithm that incorporates the ontology of knowledge graphs and partitions the knowledge graph based on classes to include more semantic information for parallel training of large-scale knowledge graph embeddings. Our preliminary results show that our algorithm performs well on several popular benchmarks.

Figures

Figures reproduced from arXiv: 2501.04613 by the authors.

Figure 1
Figure 1. Freebase class frequency analysis. After assigning a class to each entity, in order to preserve the scoring function of existing models for fact triplets, we need to partition each fact triplet by assigning it to a specific class. This is done by classifying each fact triplet based on the class of its head entity. It’s important to note that this is a preliminary, simple semantic partitioning method and requires fur… view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

35 extracted references · 25 canonical work pages

  1. [1]

    Ivana Balažević, Carl Allen, and Timothy M Hospedales. 2019. Tucker: Tensor factorization for knowledge graph completion. arXiv preprint arXiv:1901.09590 (2019)

    arXiv 2019

  2. [2]

    Jiaoyan Chen, Pan Hu, Ernesto Jimenez-Ruiz, Ole Magnus Holter, Denvar Antonyrajah, and Ian Horrocks. 2021. Owl2vec*: Embedding of owl ontologies. Machine Learning 110, 7 (2021), 1813–1845

    work page 2021

  3. [4]

    Tim Dettmers, Pasquale Minervini, Pontus Stenetorp, and Sebastian Riedel. 2018. Convolutional 2d knowledge graph embeddings. In Proceedings of the AAAI conference on artificial intelligence , Vol. 32

    work page 2018

  4. [5]

    Chi Thang Duong, Trung Dung Hoang, Hongzhi Yin, Matthias Weidlich, Quoc Viet Hung Nguyen, and Karl Aberer. 2021. Scalable robust graph embedding with Spark. Proceedings of the VLDB Endowment 15, 4 (2021), 914–922

    work page 2021

  5. [6]

    Bordes et al. 2013. Translating embeddings for modeling multi-relational data. Advances in neural information processing systems 26 (2013)

    work page 2013

  6. [7]

    Dong et al. 2014. Knowledge vault: A web-scale approach to probabilistic knowl- edge fusion. In Proceedings of the 20th ACM SIGKDD international conference on Knowledge discovery and data mining . 601–610

    work page 2014

  7. [8]

    Guo et al. 2015. Semantically smooth knowledge graph embedding. InProceedings of the 53rd Annual Meeting of the Association for Computational Linguistics and the 7th International Joint Conference on Natural Language Processing . 84–94

    work page 2015

  8. [9]

    Hao et al. 2017. An end-to-end model for question answering over knowledge base with cross-attention combining global knowledge. In Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics . 221–231

    work page 2017

Show all 35 references
  1. [10]

    Kurt et al. 2008. Freebase: a collaboratively created graph database for structur- ing human knowledge. In Proceedings of the 2008 ACM SIGMOD international conference on Management of data . 1247–1250

    2008

  2. [11]

    Lehmann et al. 2015. Dbpedia–a large-scale, multilingual knowledge base ex- tracted from wikipedia. Semantic web 6, 2 (2015), 167–195

    2015

  3. [12]

    Yang et al. 2014. Embedding entities and relations for learning and inference in knowledge bases. arXiv preprint arXiv:1412.6575 (2014)

    2014 arXiv

  4. [13]

    Zhang et al. 2016. Collaborative knowledge base embedding for recommender systems. In Proceedings of the 22nd ACM SIGKDD international conference on knowledge discovery and data mining . 353–362

    2016

  5. [14]

    Nitisha Jain, Jan-Christoph Kalo, Wolf-Tilo Balke, and Ralf Krestel. 2021. Do em- beddings actually capture knowledge graph semantics?. InThe Semantic Web: 18th International Conference, ESWC 2021, Virtual Event, June 6–10, 2021, Proceedings

    2021

  6. [15]

    Guoliang Ji, Shizhu He, Liheng Xu, Kang Liu, and Jun Zhao. 2015. Knowledge graph embedding via dynamic mapping matrix. In International joint conference on natural language processing . 687–696

    2015

  7. [16]

    Seyed Mehran Kazemi and David Poole. 2018. Simple embedding for link predic- tion in knowledge graphs. Advances in neural information processing systems 31 (2018)

    2018

  8. [17]

    Adrian Kochsiek and Rainer Gemulla. 2021. Parallel training of knowledge graph embedding models: a comparison of techniques. Proceedings of the VLDB Endowment 15, 3 (2021), 633–645

    2021

  9. [18]

    Yankai Lin, Zhiyuan Liu, Maosong Sun, Yang Liu, and Xuan Zhu. 2015. Learning entity and relation embeddings for knowledge graph completion. InTwenty-ninth AAAI conference on artificial intelligence

    2015

  10. [19]

    Changsung Moon, Paul Jones, and Nagiza F Samatova. 2017. Learning entity type embeddings for knowledge graph completion. In Proceedings of the 2017 ACM on conference on information and knowledge management . 2215–2218

    2017

  11. [20]

    Dai Quoc Nguyen, Tu Dinh Nguyen, Dat Quoc Nguyen, and Dinh Phung. 2017. A novel embedding model for knowledge base completion based on convolutional neural network. arXiv preprint arXiv:1712.02121 (2017)

    2017 arXiv

  12. [21]

    Maximilian Nickel, Volker Tresp, and Hans-Peter Kriegel. 2011. A three-way model for collective learning on multi-relational data. In Icml

    2011

  13. [22]

    Michael Schlichtkrull, Thomas N Kipf, Peter Bloem, Rianne Van Den Berg, Ivan Titov, and Max Welling. 2018. Modeling relational data with graph convolu- tional networks. In The Semantic Web: 15th International Conference, ESWC 2018, Heraklion, Crete, Greece, June 3–7, 2018, Pro...

    2018

  14. [23]

    Fabian M Suchanek, Gjergji Kasneci, and Gerhard Weikum. 2007. Yago: a core of semantic knowledge. In Proceedings of the 16th international conference on World Wide Web. 697–706

    2007

  15. [24]

    Zhiqing Sun, Zhi-Hong Deng, Jian-Yun Nie, and Jian Tang. 2019. Rotate: Knowl- edge graph embedding by relational rotation in complex space. arXiv preprint arXiv:1902.10197 (2019)

    2019 arXiv

  16. [25]

    Zequn Sun, Wei Hu, Qingheng Zhang, and Yuzhong Qu. 2018. Bootstrapping Entity Alignment with Knowledge Graph Embedding.. In IJCAI, Vol. 18. 4396– 4402

    2018

  17. [26]

    Kristina Toutanova and Danqi Chen. 2015. Observed versus latent features for knowledge base and text inference. In Proceedings of the 3rd workshop on continuous vector space models and their compositionality . 57–66

    2015

  18. [27]

    Théo Trouillon, Johannes Welbl, Sebastian Riedel, Eric Gaussier, and Guillaume Bouchard. 2016. Complex Embeddings for Simple Link Prediction. In Proceedings of The 33rd International Conference on Machine Learning . 2071–2080

    2016

  19. [28]

    Shikhar Vashishth, Soumya Sanyal, Vikram Nitin, and Partha Talukdar. 2019. Composition-based multi-relational graph convolutional networks.arXiv preprint arXiv:1911.03082 (2019)

    2019 arXiv

  20. [29]

    Denny Vrandečić and Markus Krötzsch. 2014. Wikidata: a free collaborative knowledgebase. Commun. ACM 57, 10 (2014), 78–85

    2014

  21. [30]

    Quan Wang, Zhendong Mao, Bin Wang, and Li Guo. 2017. Knowledge graph embedding: A survey of approaches and applications. IEEE Transactions on Knowledge and Data Engineering 29, 12 (2017), 2724–2743

    2017

  22. [31]

    Quan Wang, Bin Wang, and Li Guo. 2015. Knowledge base completion using embeddings and rules. In Twenty-fourth international joint conference on artificial intelligence

    2015

  23. [32]

    Zhen Wang, Jianwen Zhang, Jianlin Feng, and Zheng Chen. 2014. Knowledge graph embedding by translating on hyperplanes. In Proceedings of the AAAI conference on artificial intelligence , Vol. 28

    2014

  24. [33]

    Ruobing Xie, Zhiyuan Liu, Maosong Sun, et al. 2016. Representation learning of knowledge graphs with hierarchical types.. In IJCAI, Vol. 2016. 2965–2971

    2016

  25. [34]

    Zhanqiu Zhang, Jianyu Cai, Yongdong Zhang, and Jie Wang. 2020. Learning hierarchy-aware knowledge graph embeddings for link prediction. InProceedings of the AAAI Conference on Artificial Intelligence , Vol. 34. 3065–3072

    2020

  26. [35]

    Zhao Zhang, Fuzhen Zhuang, Meng Qu, Fen Lin, and Qing He. 2018. Knowledge graph embedding with hierarchical relation structure. In Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing . 3198–3207

    2018

  27. [36]

    Da Zheng, Xiang Song, Chao Ma, Zeyuan Tan, Zihao Ye, Jin Dong, Hao Xiong, Zheng Zhang, and George Karypis. 2020. Dgl-ke: Training knowledge graph embeddings at scale. InProceedings of the 43rd International ACM SIGIR Conference on Research and Development in Information Retrie...

    2020

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.