Pith. sign in

REVIEW 6 major objections 5 minor 20 references

Full Triple Matcher: Integrating all triple elements between heterogeneous Knowledge Graphs

T0 review · 6 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read Full Triple Matcher claims that aligning all three elements of triples across heterogeneous knowledge graphs, not just entities, is feasible and that the resulting triple-level alignments improve entity matching.

desk verdict New triple-matching task and dataset are a legitimate extension of PARIS, but the reported numbers rely on same-data threshold search and a uniqueness assumption the authors themselves show is violated. read the letter →

arxiv 2507.22914 v1 pith:HQ5H324G submitted 2025-07-20 cs.CL

Victor Eiti Yamamoto , Hideaki Takeda This is my paper · ORCID
classification cs.CL
keywords knowledgegraphsentitymatchingtripledataintegrationcontextontologyalignmentfunctionalitycompatibleanddivergenttriples
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

This paper tries to establish that knowledge graph integration should happen at the level of whole triples, not just schema or entities. It introduces Full Triple Matcher (FTM), which first aligns entity and predicate labels and then finds triple pairs that carry compatible or divergent information. The triple alignments are fed back into entity matching, so that context captured in triples improves the identification of corresponding entities. The paper argues this matters because real-world knowledge graphs differ in size, source, and information density, and existing entity-matching datasets hide those differences. FTM reports competitive entity-matching results against OAEI systems and supervised methods, and high precision and recall on a new triple-matching dataset.

What carries the argument

The load-bearing mechanism is predicate functionality, defined as the ratio of distinct subjects to the total number of triples for a predicate, together with its inverse; high values mean the predicate behaves like a function, so the object is a near-unique descriptor of the subject. Equations 5 through 7 combine entity similarity, predicate similarity, functionality, inverse functionality, and object similarity into a triple compatibility probability, and equations 11 through 13 form the divergent counterpart. Equations 14 and 15 then convert matched triples into entity similarity by taking the probability that at least one matching triple connects the entity pair and averaging that with label similarity. This bidirectional flow, from labels to triples and back to entities, is what lets contextual information improve entity matching.

What would settle it

Measure FTM on a pair of knowledge graphs in which a substantial fraction of entities have multiple objects for the same high-singularity predicate, for instance a character with two species values, and check whether triple-match precision and the entity-matching gains reported in Table 9 collapse. The paper itself reports such cases for Khan Noonien Singh, so a systematic count of multi-valued predicates and their effect on the final mappings would settle the claim.

Watch

Extended reading notes

Core claim

The central claim is that two triples from different knowledge graphs can be reliably judged as conveying the same or conflicting information, and that this triple-level judgment strengthens entity alignment. For a pair of triples $(s_1,p_1,o_1)$ and $(s_2,p_2,o_2)$, FTM multiplies the similarities of subjects, predicates, and objects by the functionality and inverse functionality of the predicates to get a compatibility probability, and uses a mirrored formula with one minus the object similarity to get a divergence probability. Matched triples are labeled compatible or divergent, and pairs of entities that appear as the subject or object of matched triples receive boosted similarity scores, averaged with label similarity. The paper reports that triple matching improves Hit@1 by 0.31 on DW-NB and 0.26 on DY-NB, datasets where labels were deliberately made unreliable, and that triple matching itself reaches F-measures around 0.79 to 0.94 across OAEI-derived test cases.

Load-bearing premise

The method assumes that each subject-predicate pair has essentially one object, so a predicate that almost always has a single object can be treated as a unique descriptor; if knowledge graphs commonly record several objects for the same subject and predicate, the triple scores and the entity alignment built on them lose reliability.

Editorial extensions

If this is right

  • Triple-level mappings give knowledge graph integration an interpretable output: each aligned pair is labeled compatible or divergent, so users can see which facts agree and which conflict.
  • Because FTM is unsupervised and statistical, it can match large knowledge graphs without training seeds, in settings where supervised embedding methods degrade as the number of entities grows.
  • In datasets where labels are manipulated or unreliable, triple matching can recover alignments that label-only matching misses, as in DW-NB and DY-NB where Hit@1 rose from about 0.6 to over 0.9.
  • Divergent triple pairs can surface factual conflicts between community-built knowledge graphs, giving a starting point for human verification or automated repair.
  • The new triple-matching dataset and task provide a way to evaluate context matching, which the paper argues existing entity-matching datasets ignore.

Reading between the lines

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

  • A natural extension is to treat multiple objects per subject-predicate pair explicitly, using outlier detection or majority voting, so that legitimate multi-valued facts like a character having two species do not become false positives.
  • If triple-level alignments are as informative as the paper suggests, similar machinery could be applied to cross-lingual knowledge graphs by replacing label matching with translation-based similarity, since the triple equations do not depend on language.
  • The divergence channel could feed a knowledge graph repair loop: instead of flagging one side as wrong, it could mark the pair for human review or for resolution by a third knowledge graph.
  • The paper's comparison suggests that statistical matchers deserve stronger weight in future entity-alignment benchmarks, which are currently dominated by supervised embedding methods.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

6 major / 5 minor

Summary. The paper introduces the Full Triple Matcher (FTM), an unsupervised method that aligns triples between heterogeneous knowledge graphs by first matching entity and predicate labels and then computing triple-level similarity from subject, predicate, object, and functionality terms. Triple pairs are classified as compatible or divergent, and the resulting triple mappings are fed back to improve entity alignment. The method is evaluated on OAEI KG-track datasets, real-size KG pairs involving DBpedia, and the DWY-NB family, and a new triple-matching benchmark is constructed from OAEI gold standards. The paper reports competitive entity matching results and high precision/recall for triple classification, and it claims that triple-level context improves entity matching when label signals are unreliable.

Significance. If the central claims hold, the paper would open a genuinely understudied task—triple-level matching with compatible/divergent labels—and show that a non-neural, unsupervised method can improve entity alignment in heterogeneous, unbalanced KG settings. The paper also ships a new triple-matching benchmark, reports runtime behavior on large KGs, and provides an explicit comparison between label-only and label-plus-triple matching (Table 9), which is a useful ablation. However, the strength of these contributions is currently limited by three load-bearing issues: the compatible/divergent definition assumes one object per subject-predicate pair even though the authors document violations of that assumption; the reported triple-matching scores use thresholds selected on the test set; and the triple benchmark is built from gold-standard entity and predicate alignments, so the reported numbers do not reflect end-to-end performance.

major comments (6)
  1. [Section 4.2, Eqs. (5)–(13); Section 6.1.2; Section 7; Table 12] The compatible/divergent classification is only coherent when each subject-predicate pair has essentially one object. Equations 5–7 multiply subject, predicate, and object similarity, while Eqs. 11–13 multiply the same factors by (1 − object similarity), so a single subject-predicate pair can be declared compatible with one object and divergent with another. The authors themselves demonstrate this: in Section 7, the species of Khan Noonien Singh is both Human and Augment in Memory Alpha, and Table 12 examples 7 and 8 label one true fact compatible and the other divergent. Because Section 6.1.2 filters predicates only by functionality above 0.8, such multi-object pairs enter the evaluation through cross-products, contaminating both the triple-level precision/recall in Table 10 and the entity bootstrap in Eq. 14. The claim that FTM accurately classifies triples as compatible or divergent is therefore not supported for KGs with multi-valued predicates; the evaluation should be restricted to predicates with functionality 1.0, or the task definition and equations should be revised to handle one-to-many cases explicitly.
  2. [Section 6.2.2; Section 7, threshold discussion] The triple-matching results in Table 10 are reported at 'best thresholds' obtained by exhaustive search over the test data, as stated in Section 6.2.2 and in the threshold discussion. This makes the reported precision, recall, and F-measure optimistic estimates of achievable performance and does not support the claim of high accuracy on unseen data. The authors should either use a validation split to select thresholds or report threshold-averaged curves (as in Figure 4) across all datasets, and should state explicitly which numbers are post-hoc selections rather than predictions.
  3. [Section 6.1.2; Table 10] The triple-matching benchmark is constructed by taking gold-standard entity and predicate alignments and then extracting all triples whose subjects and predicates appear in those alignments. Consequently, the triple-matching evaluation in Table 10 assumes perfect entity and predicate alignments; it does not measure how triple matching performs when the upstream label matching and entity alignment contain the errors that occur in actual use. The paper should report an end-to-end version of triple matching (using FTM's own entity and predicate mappings) and compare it with the oracle-alignment version, or should state clearly that Table 10 measures only the triple-scoring component under oracle alignments.
  4. [Table 8; Section 6.2.1] The comparison with supervised methods reuses Hit@k numbers from Zhang et al. (2022) and Zhang et al. (2023) without controlling for training seeds, dataset versions, or hardware, and the paper presents FTM as outperforming supervised methods across all datasets. This is a strong comparative claim that requires either reproducing those methods under the same evaluation protocol or softening the conclusion to 'comparable to published numbers' with the appropriate caveats. The point is secondary to the triple-matching contribution but is load-bearing for the paper's state-of-the-art claim.
  5. [Section 4.2, Eq. (15) and Section 5.2.1] The co-recursive dependence between triple similarity (Eqs. 5–7) and entity similarity (Eq. 14) is handled by averaging with label similarity and by defaulting missing triple similarity to 0.5. This is a legitimate way to break the circularity, but the default value of 0.5 and the equal weighting in Eq. 15 are free parameters whose influence on the final results is not analyzed. The paper should provide a sensitivity analysis for these parameters, particularly since the improvement in Table 9 is driven by exactly the regime where triple similarity and label similarity disagree.
  6. [Equation 8 in Section 4.2; Algorithm 1 lines 12–13] The example calculation is internally inconsistent and should be corrected: the text after Eq. 7 says the functionality result is 0.16, but Eq. 8 computes 0.29 and Eq. 10 uses 0.29. Also, in Algorithm 1, line 13 defines T′2 using T1 instead of T2, and line 12 retrieves triples where the entity is the object, which is then used in the 'inbound' step that should retrieve triples where the entity is the subject. These are local presentation errors, but they make the algorithm description difficult to follow and should be fixed before publication.
minor comments (5)
  1. [Section 4.2, Eq. (9)] Equation 9 appears to use the inverse functionality values 0.16 and 0.20, but the surrounding text reports inverse functionality of 0.16 for dbo:located and 0.20 for mcu:located, so the numerical example is consistent; the discrepancy with the '0.16' in the prose remains and should be resolved.
  2. [Section 5.1, Table 2] The label matching thresholds (1.0, 0.9, 0.8, 0.7) are presented without justification or sensitivity analysis; a brief sentence on how these values were chosen would improve reproducibility.
  3. [Section 6.1.2, 'triple matching' metric] The definition of true positive, false positive, and false negative for triple matching is stated only in words; a formal definition analogous to Eqs. 26–28 would prevent ambiguity about whether a false positive is a gold-standard pair that is misclassified or a non-gold pair that is predicted.
  4. [Table 11 and Figure 4] The confusion matrix in Table 11 is useful, but the marginal numbers should be checked for arithmetic consistency; the text would benefit from a sentence explaining the relation between the compatible and divergent confusion matrices.
  5. [Section 7, final paragraph] The phrase 'we do not infer which one is the correct one' is important and should appear earlier in the paper, ideally in the introduction or problem definition, so that readers do not interpret 'divergent' as 'false'.

Circularity Check

1 steps flagged · score 2.0 of 10

No derivation circularity in the co-recursive entity/triple loop; minor evaluation-level circularity in the triple-matching gold standard, which is constructed from the same entity/predicate gold standard and similar object-similarity rules used by the model.

  1. other [Section 6.1.2 (Triple matching) and Table 10]
    "First, we selected the predicates in the gold standard with functionality above 0.8. Second, we extracted all triples with entities in the gold standard and predicates in the filtered set. Finally, we created the list of mappings between triples with corresponding entity and predicate pairs in the gold standard. In the manual evaluation, if the objects are both entities, we define them as compatible if they are in the gold standard, divergent if a mapping exists to another entity."

    The triple-matching gold standard is not an independent ground truth: it is generated from the same OAEI entity/predicate gold standard that FTM is trying to reproduce, and the compatible/divergent labels are assigned by object-comparison heuristics (numeric tolerance, string/type similarity) that mirror the object-similarity functions in Table 1 and the (1 - Pr_obj) divergence definition in Eqs. 11-13. Reporting precision/recall/F-measure against this self-derived labeling therefore partly measures agreement with the label-construction rule rather than an externally validated semantic distinction. The entity-matching results remain externally benchmarked, so this is an evaluation-level circularity, not a tautological derivation of the model equations.

full rationale

The central co-recursive loop (Eqs. 5-7 compute triple similarity from entity similarity; Eq. 14 computes entity similarity from triple similarity; Eq. 15 averages with label similarity) is anchored by an external signal, label matching via strings and BERT, so it is not equivalent to its inputs by construction. The improvements on DW-NB and DY-NB come from exact common-literal matching, an independent mechanism. No load-bearing self-citations appear; PARIS is an external published method, and the paper does not rely on a self-cited uniqueness theorem or ansatz. The only notable circularity concern is the triple-matching evaluation: the gold standard is built from the same OAEI entity/predicate mappings the method targets, and the manual compatible/divergent rules use similarity criteria similar to the model's object-similarity table. This inflates the reported triple-matching F-measures to some degree but does not make the derivation itself circular. Score 2 reflects this minor evaluation-level self-confirmation rather than a fundamental reduction of the claimed predictions to fitted inputs.

Assumptions & free parameters 7 free parameters · 7 assumptions · 0 invented entities

The model's load-bearing free parameters are hand-chosen label similarity weights, a default 0.5, an equal 0.5/0.5 combination, per-dataset thresholds obtained by exhaustive search, and algorithmic caps. The axioms are assumptions about functionality as informativeness, uniqueness of subject-predicate pairs, the independence noise model, and open-world gold standards. No new entities such as particles, forces, or mediators are introduced.

free parameters (7)
  • Label similarity tier weights = 1.0, 0.9, 0.8, 0.7, 0.7 x fuzzy, 0.7 x BERT
    Hand-chosen constants in Table 2 that fix all label-level confidences before triple matching; they propagate through Equations 15 and 5 to 7.
  • Default missing triple similarity = 0.5
    Section 4.2 states that when an entity pair lacks label or triple similarity, a default value of 0.5 is used for triple similarity. This neutral prior affects all cold-start similarity scores.
  • Label/triple combination weight = 0.5
    Equation 15 and Algorithm 1 line 8 combine label and triple similarity with equal weight, with no calibration or justification of that balance.
  • Entity matching decision thresholds per dataset = 0.90 to 0.94
    Section 6.2.1 states thresholds were determined by exhaustively evaluating all possible values to maximize F-measure, making the reported F-measure values fitted rather than predicted.
  • Triple matching thresholds per dataset = 0.11 to 0.74 for compatible; 0.14 to 0.45 for divergent
    Table 10 reports thresholds selected to maximize F-measure on the same datasets, so the precision and recall numbers are tuned on the evaluation set.
  • Functionality cutoff for triple dataset construction = 0.8
    Section 6.1.2 selects predicates with functionality above 0.8 to build the triple matching benchmark, restricting the evaluation to a subset of predicates.
  • Candidate and iteration caps = top-10 pairs, max 10 iterations, 10 percent early-stop
    Sections 5.2.3, 5.2.4 and Algorithm 1 use arbitrary caps on candidate pairs and iterations; results may depend on these choices.
assumptions (7)
  • domain assumption Functionality and inverse functionality are valid proxies for how informative a predicate is for entity and triple matching.
    Section 4.2 asserts that predicates with high functionality or inverse functionality better represent entity characteristics. This is inherited from PARIS and assumed, not demonstrated for heterogeneous KGs.
  • domain assumption Two triples are similar only if subjects, predicates, and objects are all similar, and the similarity factors combine by multiplication.
    Equations 5 and 6 define triple probability as a product of entity, predicate, object similarities and functionality; no independence or calibration argument is given.
  • standard math The noise model 1 minus the product of (1 minus probability) for combining independent evidence is applicable here.
    Used in Equations 7, 13, 14, and 23 to 25 following PARIS; it is a standard probabilistic combination but requires an independence assumption that is not tested.
  • domain assumption Label similarity between entities is an independent and reliable anchor for the recursive entity-triple equations.
    Equation 15 averages label and triple similarity, and Algorithm 1 uses label mappings as the initial M_e; if labels are missing or misleading, the recursion has no external grounding.
  • domain assumption Each entity-predicate pair has at most one object, so high-functionality predicates describe entities uniquely.
    Section 7 acknowledges exceptions such as Khan having both Human and Augment species, and future work plans to go beyond 1-1 entity-predicate triples. The cross-product matching relies on this uniqueness.
  • domain assumption The OAEI gold standard is sufficiently complete for open-world evaluation, and unlabeled pairs can be ignored.
    Section 6.1.1 defines precision and recall only over gold-standard pairs under the open-world assumption; if the gold standard misses correct matches, reported precision is not end-to-end.
  • domain assumption Object similarity heuristics in Table 1 correspond to semantic compatibility.
    Numeric tolerance, date conversion, and fuzzy string methods are chosen by the authors, not derived or validated externally.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Full Triple Matcher: Integrating all triple elements between heterogeneous Knowledge Graphs." pith.science (2026). https://pith.science/paper/HQ5H324G

@misc{pith2026250722914,
  author       = {Pith},
  title        = {Pith review of: Full Triple Matcher: Integrating all triple elements between heterogeneous Knowledge Graphs},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/HQ5H324G}},
  note         = {Machine review of arXiv:2507.22914}
}
read the original abstract

Knowledge graphs (KGs) are powerful tools for representing and reasoning over structured information. Their main components include schema, identity, and context. While schema and identity matching are well-established in ontology and entity matching research, context matching remains largely unexplored. This is particularly important because real-world KGs often vary significantly in source, size, and information density - factors not typically represented in the datasets on which current entity matching methods are evaluated. As a result, existing approaches may fall short in scenarios where diverse and complex contexts need to be integrated. To address this gap, we propose a novel KG integration method consisting of label matching and triple matching. We use string manipulation, fuzzy matching, and vector similarity techniques to align entity and predicate labels. Next, we identify mappings between triples that convey comparable information, using these mappings to improve entity-matching accuracy. Our approach demonstrates competitive performance compared to leading systems in the OAEI competition and against supervised methods, achieving high accuracy across diverse test cases. Additionally, we introduce a new dataset derived from the benchmark dataset to evaluate the triple-matching step more comprehensively.

Figures

Figures reproduced from arXiv: 2507.22914 by the authors.

Figure 1
Figure 1. Triple example with high functionality predicate [PITH_FULL_IMAGE:figures/full_fig_p010_1.png] view at source ↗
Figure 2
Figure 2. Triple example with low functionality predicate [PITH_FULL_IMAGE:figures/full_fig_p011_2.png] view at source ↗
Figure 3
Figure 3. Full Triple Matcher Flowchart point to an entity in one KG, while in another, it may reference a literal such as a string representing a name. From an algorithmic perspective, our model supports multiple input formats, including TTL, RDF/XML, and SPARQL endpoints. The latter is particularly useful for large-scale knowledge graphs. In contrast, PARIS is limited to the N-Triples format and maintains all potential enti… view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Triple matching threshold evolution strings with relative similarity, often resulting in a moderate score for compatibility and divergence probabilities [PITH_FULL_IMAGE:figures/full_fig_p026_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

20 extracted references · 15 canonical work pages

  1. [6]

    Dbkwik: A consolidated knowledge graph from thousands of wikis

    “Dbkwik: A consolidated knowledge graph from thousands of wikis. ” In:2018 IEEE International Conference on Big Knowledge (ICBK) . IEEE, 17–24. Sven Hertling and Heiko Paulheim. 2020a. “Dbkwik: extracting and integrating knowledge from thousands of wikis. ” Knowledge and Information Systems , 62, 6, 2169–2190. Sven Hertling and Heiko Paulheim

    work page 2018

  2. [8]

    Deep Active Alignment of Knowledge Graph Entities and Schemata

    “Deep Active Alignment of Knowledge Graph Entities and Schemata. ” Proc. ACM Manag. Data , 1, 2, Article 159, (June 2023), 26 pages. doi: 10.1145/3589304. Peng Huang, Meihui Zhang, Ziyue Zhong, Chengliang Chai, and Ju Fan

    doi:10.1145/3589304 2023

  3. [9]

    Representation Learning for Entity Alignment in Knowledge Graph: A Design Space Exploration

    “Representation Learning for Entity Alignment in Knowledge Graph: A Design Space Exploration. ” In: 2024 IEEE 40th International Conference on Data Engineering (ICDE). IEEE, 3462–3475. Xuhui Jiang, Chengjin Xu, Yinghan Shen, Fenglong Su, Yuanzhuo Wang, Fei Sun, Zixuan Li, and Huawei Shen

    work page 2024

  4. [10]

    Rethinking GNN-based Entity Alignment on Heterogeneous Knowledge Graphs: New Datasets and A New Method

    “Rethinking GNN-based Entity Alignment on Heterogeneous Knowledge Graphs: New Datasets and A New Method. ” arXiv preprint arXiv:2304.03468. Ernesto Jiménez-Ruiz and Bernardo Cuenca Grau

    arXiv

  5. [13]

    A Critical Re-Evaluation of Neural Methods for Entity Alignment

    “A Critical Re-Evaluation of Neural Methods for Entity Alignment. ”Proc. VLDB Endow., 15, 8, (Apr. 2022), 1712–1725. doi: 10.14778/3529337.3529355. Shuangyan Liu, Mathieu d’Aquin, and Enrico Motta

    work page arXiv 2022

  6. [14]

    Attribute Enhancement using Aligned Entities between Knowledge Graphs

    “Attribute Enhancement using Aligned Entities between Knowledge Graphs. ” In:2023 IEEE 17th International Conference on Semantic Computing (ICSC) , 191–198. doi: 10.1109/ICSC56153.202 3.00038. ACM Trans. Web, Vol. 37, No. 4, Article

    work page doi:10.1109/icsc56153.202 2023

  7. [16]

    PARIS: Probabilistic Alignment of Relations, Instances, and Schema

    “PARIS: Probabilistic Alignment of Relations, Instances, and Schema. ”Proc. VLDB Endow., 5, 3, (Nov. 2011), 157–168. doi: 10.14778/2078331.2078332. Zequn Sun, Qingheng Zhang, Wei Hu, Chengming Wang, Muhao Chen, Farahnaz Akrami, and Chengkai Li

    work page arXiv 2011

  8. [18]

    Two heads are better than one: Integrating knowledge from knowledge graphs and large language models for entity alignment

    “Two heads are better than one: Integrating knowledge from knowledge graphs and large language models for entity alignment. ”arXiv preprint arXiv:2401.16960. Rui Zhang, Yixin Su, Bayu Distiawan Trisedya, Xiaoyan Zhao, Min Yang, Hong Cheng, and Jianzhong Qi

    arXiv

Show all 20 references
  1. [19]

    A benchmark and comprehensive survey on knowledge graph entity alignment via representation learning

    “A benchmark and comprehensive survey on knowledge graph entity alignment via representation learning. ”The VLDB Journal, 31, 5, 1143–1168. Received 20 February 2007; revised 12 March 2009; accepted 5 June 2009 ACM Trans. Web, Vol. 37, No. 4, Article

    2007

  2. [111]

    Publication date: August 2018

    2018

  3. [162]

    https://proceedings.mlr.press/v16 2/guo22i.html

    PMLR, (July 2022), 8145–8156. https://proceedings.mlr.press/v16 2/guo22i.html. Sven Hertling and Heiko Paulheim

    2022

  4. [2004]

    NLTK: The Natural Language Toolkit

    “NLTK: The Natural Language Toolkit. ” In:Proceedings of the ACL Interactive Poster and Demonstration Sessions . Association for Computational Linguistics, Barcelona, Spain, (July 2004), 214–217. https://aclanthology.org/P04-3031. Antoine Bordes, Nicolas Usunier, Alberto Garci...

    2004

  5. [2009]

    Metric functional dependencies

    “Metric functional dependencies. ” In: 2009 IEEE 25th International Conference on Data Engineering . IEEE, 1275–1278. Manuel Leone, Stefano Huber, Akhil Arora, Alberto García-Durán, and Robert West. Apr

    2009

  6. [2011]

    Logmap: Logic-based and scalable ontology matching

    “Logmap: Logic-based and scalable ontology matching. ” In:The Semantic Web–ISWC 2011: 10th International Semantic Web Conference, Bonn, Germany, October 23-27, 2011, Proceedings, Part I

    2011

  7. [2016]

    SPrank: Semantic Path-Based Ranking for Top-N Recommendations Using Linked Open Data

    “SPrank: Semantic Path-Based Ranking for Top-N Recommendations Using Linked Open Data. ”ACM Trans. Intell. Syst. Technol., 8, 1, Article 9, (Sept. 2016), 34 pages. doi: 10.1145/2899005. Mina Abd Nikooie Pour et al

    2016 doi

  8. [2018]

    BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding

    “BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding. ”CoRR, abs/1810.04805. http://arxiv.org/abs/1810.04805 arXiv: 1810.04805. Jianyong Fang and Xuefeng Yan

    arXiv

  9. [2020]

    Neighborhood Matching Network for Entity Alignment

    “Neighborhood Matching Network for Entity Alignment. ” In:Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics . Ed. by Dan Jurafsky, Joyce Chai, Natalie Schluter, and Joel Tetreault. Association for Computational Linguistics, Online, (July 2...

    2020 doi

  10. [2022]

    Gollum: A gold standard for large scale multi source knowledge graph matching

    “Gollum: A gold standard for large scale multi source knowledge graph matching. ” arXiv preprint arXiv:2209.07479. Sven Hertling and Heiko Paulheim. 2020b. “The knowledge graph track at OAEI: Gold standards, baselines, and the golden hammer bias. ” In:The Semantic Web: 17th In...

    2020 arXiv

  11. [2023]

    Accessed August 11, 2023, https://github.com/maxbachmann/RapidFuzz

    Rapid fuzzy string matching in Python and C++ using the Levenshtein Distance . Accessed August 11, 2023, https://github.com/maxbachmann/RapidFuzz. (2023). ACM Trans. Web, Vol. 37, No. 4, Article

    2023

  12. [2024]

    LLM-Align: Utilizing Large Language Models for Entity Alignment in Knowledge Graphs

    “LLM-Align: Utilizing Large Language Models for Entity Alignment in Knowledge Graphs. ”arXiv preprint arXiv:2412.04690. Isabel F Cruz, Flavio Palandri Antonelli, and Cosmin Stroe

    arXiv

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.