Pith. sign in

REVIEW 4 major objections 5 minor 68 references

VITA: Versatile Time Representation Learning for Temporal Hyper-Relational Knowledge Graphs

T0 review · 4 major / 5 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read A single time triplet $(c,t_1,t_2)$ represents since, until, period, and invariant fact validity, and VITA learns from time value plus timespan to outperform baselines by up to 75.3%.

desk verdict A clean time-triplet representation worth publishing, but the empirical advantage is untestable without code/data and the headline 75.3% is an average over a single baseline on custom data. read the letter →

arxiv 2505.11803 v1 pith:KNZQTDDV submitted 2025-05-17 cs.AI cs.SC

ChongIn Un , Yuhuan Lu , Tianyue Yang , Dingqi Yang This is my paper · ORCID
classification cs.AIcs.SC
keywords temporalknowledgegraphlinkpredictionhyper-relationaltimerepresentationvalidityembeddingvalueencodernumericliterals
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 argues that existing ways of attaching time to knowledge-graph facts—temporal snapshots, granularity-based expansion, or treating timestamps as ordinary numeric qualifiers—either depend on ad-hoc interval choices or discard facts with infinite validity. It proposes a versatile time triplet $(c,t_1,t_2)$ with $c\in\{\mathrm{Since},\mathrm{Until},\mathrm{Period},\mathrm{Invariant}\}$ and $t_1,t_2$ real numbers or $\pm\infty$ tokens, claimed to cover exactly the four kinds of temporal validity a fact can have. Building on this representation, VITA is an encoder–decoder model that learns time values and validity timespan separately and, under one masked training scheme, predicts missing entities, relations, time values, and numeric literals. The paper reports that VITA outperforms the best applicable baselines by 4.4–8.7% in entity prediction and by up to 75.3% in time prediction, with ablations showing the representation itself, not the architecture alone, drives much of the gain.

What carries the argument

The load-bearing object is the versatile time triplet $(c,t_1,t_2)$, whose conjunction $c$ selects one of four validity types and whose values are either real numbers or learned $\pm\infty$ tokens; completeness comes from allowing open-ended intervals instead of only closed periods. Two mechanisms carry the learning: the Time Value Encoder (TVE), which maps real-valued times through a min-max-normalized unit embedding and reserves dedicated embedding vectors for $\pm\infty$, and the TimeSpan Fuser (TSF), a translation-invariant learnable sinusoidal encoding, extended to infinity tokens, that injects the validity span into base-triplet and qualifier features. The design choice that prevents information leakage is that TSF features are fused into the entity and qualifier decoders but not into the time-triplet decoder, so predicting $t_1$ or $t_2$ must rely on context rather than on a transformed copy of the answer. These mechanisms together let a single masked training pass supervise all prediction heads from one factorized representation of the fact.

What would settle it

A reader could settle the claim by sampling facts directly from a public knowledge graph—open-ended start times, open-ended end times, and no time qualifiers—without the paper's merge-and-map pipeline, and comparing VITA's time-prediction error against the best qualifier-based baseline; if the reported 75.3% advantage shrinks to a few percent once the baseline is also given infinity handling, the versatile representation is not what carries the result.

Watch

Extended reading notes

Core claim

The central discovery is representational: every temporal validity of a fact can be written exactly as a time triplet $(c,t_1,t_2)$, where the conjunction $c$ encodes the validity type—valid since $t_1$, valid until $t_2$, valid in the period $[t_1,t_2]$, or always valid—and the two time values may be real numbers or special $\pm\infty$ tokens. This turns a temporal hyper-relational fact into $\{(s,r,o),(c,t_1,t_2),\{(k_i,v_i)\}\}$ and removes the need to choose a snapshot interval or a timestamp granularity, because the fact is stored once at whatever precision the data has. The paper argues that the gains come from treating the time triplet as a first-class component: VITA separates the base triplet, time triplet, and qualifiers into three encoders, lets self-attention exchange information among them, and feeds the validity timespan into entity and qualifier prediction while deliberately withholding it from time prediction to avoid leaking the answer. The claimed result is that this design beats the best-performing baselines in all four link-prediction tasks, with the largest margin in predicting time values.

Load-bearing premise

The load-bearing premise, introduced with the dataset construction, is that the three custom benchmark datasets—built by crawling qualifiers and time information from public knowledge graphs and manually mapping relation identifiers, with about 0.2% of unmappable facts dropped—are fair and unbiased for comparing VITA against models designed for other time representations; if the construction favours VITA's triplet, the reported gains would not transfer.

Editorial extensions

If this is right

  • Temporal link prediction no longer needs a hand-chosen snapshot interval or granularity; one representation serves facts of any precision and any validity length.
  • Facts that are valid forever, or have only a start or only an end, become ordinary training examples instead of being dropped or artificially truncated, because $\pm\infty$ are explicit tokens.
  • A single masked model answers entity, relation, time, and numeric-literal queries from one training pass, so applications need one pipeline rather than separate specialists.
  • Storing each fact once instead of expanding it over its validity removes the oversampling bias and the out-of-memory failures that long-validity facts cause in expansion-style baselines.
  • The large reported margin in time prediction indicates that validity timespan is a strong learnable signal, not a passive annotation.

Reading between the lines

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

  • The paper leaves implicit that the time triplet is a representation layer, not a model; any hyper-relational encoder that adopts the triplet and infinity-token handling could plausibly inherit part of the reported advantage, independent of VITA's specific transformer blocks.
  • A testable extension: the benchmarks are dominated by Period facts (about 83–100%), with almost no Invariant cases on two of them, so a dataset where Since, Until, and Invariant are the majority would directly stress-test the completeness claim that the current experiments cannot fully confirm.
  • VITA is evaluated only in the interpolation setting (missing elements at historical timestamps); adapting the triplet representation to extrapolation, i.e., forecasting links after the training window, is a natural next step the paper does not pursue.
  • Treating open-ended validity as a token rather than a missing value suggests a model-agnostic normalization idea for temporal knowledge graphs: other numeric qualifiers that can be undefined or infinite might benefit from the same token-versus-value handling.
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. This paper proposes a versatile time representation for temporal hyper-relational knowledge graphs, encoding the temporal validity of a fact as a time triplet (c, t1, t2) with c in {Since, Until, Period, Invariant}, which covers the four common single-interval validity types including unbounded intervals. It then introduces VITA, an encoder-decoder model with separate encoders for the base triplet, the time triplet, and the qualifiers, a Time Value Encoder to handle both real-valued and infinite time values, and a TimeSpan Fuser that injects validity-span features into the base-triplet and qualifier decoders. The model is trained with a masked strategy and evaluated on entity, relation, time, and numeric-literal prediction over four datasets (Wiki, YAGO, wikipeople, ICEWS14), reporting consistent gains over a substantial set of TKG and HKG baselines, with the largest advantages in time prediction relative to HyNT. Ablation studies that remove the time triplet and the TimeSpan Fuser support the key design choices.

Significance. The central claim—that explicitly modeling the type and span of temporal validity, rather than treating time as an ordinary numeric qualifier, improves link prediction—is well-motivated and the proposed representation is simple and intuitively appealing. The paper's strengths include a broad comparison against state-of-the-art TKG and HKG baselines, consistent improvements across four datasets, and ablations that isolate the contributions of the time triplet and TimeSpan Fuser. If the results are reproducible, the work offers a useful building block for temporal hyper-relational KG reasoning. However, the empirical support depends on custom datasets whose construction may advantage VITA, the model's own hyperparameters are not reported, and code and data are not released, so the reported gains cannot currently be independently verified.

major comments (4)
  1. [§5.1.1] The dataset-construction procedure for YAGO and Wiki may give VITA an information advantage over HyNT. The text states that for YAGO "a few collected qualifiers here also contain time information, we update the time information of the corresponding facts according to the (latest) qualifiers we collected from Wikidata," but it does not state whether these time-valued qualifiers are subsequently removed from the qualifier set. If they remain, VITA receives time information through both the time triplet and the qualifier encoder, while HyNT receives it only as qualifiers; the reported 75.3% average improvement in time prediction could then be partly due to redundant input rather than the representation. The authors should clarify the removal policy and report an ablation in which time-valued qualifiers are stripped from the qualifier set for both VITA and HyNT.
  2. [Appendix A / §5.1.2] VITA's own hyperparameters are never reported. The appendix provides hyperparameter settings for all baselines but not for VITA (embedding size, number of self-attention layers and heads, learning rate, batch size, dropout, training epochs, and the lambda value per dataset). Combined with the absence of code and dataset release, this makes the experimental results unreproducible. Please provide a complete hyperparameter table and release the datasets and code, or at minimum make the dataset-construction code and exact train/validation/test splits available.
  3. [Table 4 / §5.2.3] The time and numeric-literal prediction comparison appears to lack a common protocol. The paper does not describe how HyNT is configured for time prediction (for example, whether it predicts a softmax distribution over discretized numeric tokens or a real-valued regression) nor how the reported MSE is computed relative to the normalized time values used in TVE. A mismatch in the prediction target or evaluation scaling would make the large reported improvement (75.3% average relative MSE reduction) difficult to interpret. Please specify the exact prediction and evaluation procedure for each baseline on these tasks, including normalization and mapping back to the original time scale.
  4. [Abstract / §5.2.3] The claim of "up to 75.3%" improvement is ambiguous and arguably inaccurate. From Table 4, the relative time-MSE improvements over HyNT are about 69.6% on Wiki, 65.5% on YAGO, 67.1% on wikipeople, and 98.9% on ICEWS14; thus 75.3% is the average improvement across the four datasets, not the maximum. Please report per-dataset improvements and adjust the wording to "average improvement of 75.3%" or state the actual maximum, so the abstract is not misleading.
minor comments (5)
  1. [§3, Definition 3.1] The "Completeness" property is overstated. A single time triplet (c, t1, t2) represents one contiguous validity interval, possibly unbounded; facts with multiple disjoint validity intervals (for example, two non-consecutive terms in office) cannot be represented. Please qualify the completeness claim by specifying that it holds for a single validity interval per fact.
  2. [§4.3] The paper states that infinity tokens are not predicted because the conjunction uniquely determines them, but Definition 3.3 includes t1 and t2 as possible missing elements. Please clarify whether test queries with missing infinity tokens are excluded from the time-prediction evaluation and how this affects the reported MSE.
  3. [§5.1.2] The time granularity for expansion-based baselines is set to 1 year, but ICEWS14 originally has daily timestamps. Please confirm that 1 year is the finest granularity used and discuss the sensitivity of the results to this choice, since coarser pooling could disadvantage baselines on short-duration events.
  4. [Table 3] On Wiki, GRAN achieves a higher MRR (0.9861) than VITA (0.9856), and on YAGO HyNT is higher (0.9483 vs 0.9462), yet the text describes VITA as among the best-performing. Please clarify the ranking or add a footnote stating that VITA is best on average but not on every dataset.
  5. [Throughout] There are several typographical and consistency issues: "TKHG" in Definition 3.3 should be "THKG"; "TSE" in Section 4.2 should likely be "TSF"; the spelling of "TeRo/TeRO" is inconsistent; and the reference numbering in the related-work text appears duplicated for [62]. A careful proofread would improve presentation.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: versatile time representation is a definitional encoding scheme, and the claimed gains rest on standard held-out supervised evaluation.

full rationale

I walked the derivation chain from Definition 3.1 through the VITA architecture (Section 4) and the experiments (Section 5). The versatile time triplet (c, t1, t2) with c in {Since, Until, Period, Invariant} is introduced as a representational scheme; its coverage of the four temporal-validity types is true by construction, but the paper does not present that coverage as a derived prediction. The load-bearing empirical claims (up to 75.3% gains in link prediction) come from a standard train/validation/test split on masked-position prediction tasks, with no test-set fitting and no parameter fitted to the reported results. The time-value encoder (TVE) and time decoder use learned linear regressions on min-max normalized values, and the masked time value is removed and must be inferred from the remaining context, so time prediction is not identical to an input by construction. The only author-overlap citation is HINGE [42], used as a baseline and as general motivation for separating the base triplet from qualifiers; it is not invoked to justify VITA's time representation or to forbid alternative designs, and it is not a uniqueness theorem. Dataset construction in Section 5.1.1 (merging timestamps, crawling qualifiers, manual YAGO-to-Wikidata mapping) raises fairness and reproducibility concerns rather than circularity; notably, VITA's own hyperparameters are not reported even though Appendix A gives baseline settings, and the paper does not state whether time-valued qualifiers are removed from the qualifier set after being promoted to the time triplet. These are correctness-risk issues, not circular reductions. I find no step where a reported prediction reduces to a fitted input, a self-citation chain, or a definitional relabeling.

Assumptions & free parameters 4 free parameters · 5 assumptions · 0 invented entities

VITA is an empirical ML paper, so the main 'free parameters' are the model's hyperparameters and data-derived normalization bounds, which are fitted to the training data. No new physical entities are introduced. The key unstated assumptions are about the completeness of the four-type time representation and about the fairness of the author-constructed benchmarks.

free parameters (4)
  • lambda (loss balancing weight) = 1 on most datasets
    Eq. 5 balances cross-entropy and MSE losses; chosen empirically, aligning with HyNT.
  • VITA model hyperparameters (embedding size, learning rate, layers, heads) = not reported
    The paper gives baseline hyperparameters in Appendix A but does not report VITA's own settings; the reported results depend on these choices.
  • min-max normalization bounds for time values and numeric literals = data-dependent
    Time Value Encoder and relation-specific normalization phi_r use min-max scaling over training data; these bounds are fitted to the dataset.
  • learnable time encoding parameters (omega, phi) in TimeSpan Fuser = learned during training
    Eq. 2 defines learnable angular frequencies and phases; these are fitted to data and central to the timespan feature.
assumptions (5)
  • domain assumption Four time conjunctions (Since, Until, Period, Invariant) are sufficient to represent all temporal validity of facts in the benchmark datasets.
    Definition 3.1 assumes completeness; the datasets are constructed to fit these four types, but real temporal facts may have more nuanced expressions.
  • domain assumption Merging facts that share the same triplet and qualifiers but have different start/end times into a single Period fact preserves the information needed for link prediction.
    Section 5.1.1 describes this merging for Wiki, YAGO, and ICEWS14; if merging discards useful temporal distinctions, the benchmark is altered.
  • domain assumption Manual mapping of YAGO relations to Wikidata and dropping unmappable facts (about 0.2%) does not introduce bias favoring VITA.
    Section 5.1.1, YAGO description.
  • domain assumption The interpolation setting means test queries come from the same time range as training, so regression on normalized time values is appropriate.
    Section 3 states the paper focuses on interpolation.
  • standard math Transformer self-attention and the translation-invariant time encoding from Xu et al. function as described in the cited literature.
    Eq. 2 reuses the TGAT time encoding.

how reviews work

0 comments
Cite this review

Pith. "Pith review of VITA: Versatile Time Representation Learning for Temporal Hyper-Relational Knowledge Graphs." pith.science (2026). https://pith.science/paper/KNZQTDDV

@misc{pith2026250511803,
  author       = {Pith},
  title        = {Pith review of: VITA: Versatile Time Representation Learning for Temporal Hyper-Relational Knowledge Graphs},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/KNZQTDDV}},
  note         = {Machine review of arXiv:2505.11803}
}
abstract

Knowledge graphs (KGs) have become an effective paradigm for managing real-world facts, which are not only complex but also dynamically evolve over time. The temporal validity of facts often serves as a strong clue in downstream link prediction tasks, which predicts a missing element in a fact. Traditional link prediction techniques on temporal KGs either consider a sequence of temporal snapshots of KGs with an ad-hoc defined time interval or expand a temporal fact over its validity period under a predefined time granularity; these approaches not only suffer from the sensitivity of the selection of time interval/granularity, but also face the computational challenges when handling facts with long (even infinite) validity. Although the recent hyper-relational KGs represent the temporal validity of a fact as qualifiers describing the fact, it is still suboptimal due to its ignorance of the infinite validity of some facts and the insufficient information encoded from the qualifiers about the temporal validity. Against this background, we propose VITA, a $\underline{V}$ersatile t$\underline{I}$me represen$\underline{TA}$tion learning method for temporal hyper-relational knowledge graphs. We first propose a versatile time representation that can flexibly accommodate all four types of temporal validity of facts (i.e., since, until, period, time-invariant), and then design VITA to effectively learn the time information in both aspects of time value and timespan to boost the link prediction performance. We conduct a thorough evaluation of VITA compared to a sizable collection of baselines on real-world KG datasets. Results show that VITA outperforms the best-performing baselines in various link prediction tasks (predicting missing entities, relations, time, and other numeric literals) by up to 75.3%. Ablation studies and a case study also support our key design choices.

Figures

Figures reproduced from arXiv: 2505.11803 by the authors.

Figure 1
Figure 1. Different representation schemes for temporal [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Four types of temporal validity represented by the [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Overview of VITA. Definition 3.3 (Link Prediction over THKG). The link predic￾tion over TKHG is to predict a missing element in a temporal hyper￾relational fact {(𝑠, 𝑟, 𝑜), (𝑐, 𝑡1, 𝑡2), {(𝑘𝑖 , 𝑣𝑖)}}. The missing element could be any element 𝑠, 𝑟, 𝑜, 𝑐, 𝑡1, 𝑡2, 𝑘𝑖 , or 𝑣𝑖 in this fact. Depending on the position and modality of the missing element, link prediction tasks can be classified into 1) entity prediction wher… view at source ↗
Figures from the paper (2 more)
Figure 1
Figure 1. Figure 1: For other methods in this category, we remove the time in [PITH_FULL_IMAGE:figures/full_fig_p006_1.png]
Figure 4
Figure 4. Figure 4: Case Study. versatile time representation) captures this subtle clue and ranks André Cruz in front of Léo Lima, while w/o VTR (representing time information as qualifiers) fails in this case. 6 Conclusion In this paper, we introduce VITA, a Versatile tIme represenTAtio…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

68 extracted references · 37 canonical work pages

  1. [1]

    Ralph Abboud, Ismail Ceylan, Thomas Lukasiewicz, and Tommaso Salvatori

    work page

  2. [2]

    Antoine Bordes, Nicolas Usunier, Alberto Garcia-Duran, Jason Weston, and Ok- sana Yakhnenko. 2013. Translating embeddings for modeling multi-relational data.Advances in neural information processing systems26 (2013)

    2013

  3. [3]

    Dan Brickley, Ramanathan V Guha, and Brian McBride. 2014. RDF Schema 1.1. W3C recommendation25 (2014), 2004–2014

    work page 2014

  4. [4]

    Li Cai, Xin Mao, Yuhao Zhou, Zhaoguang Long, Changxu Wu, and Man Lan

  5. [5]

    Kai Chen, Ye Wang, Yitong Li, Aiping Li, Han Yu, and Xin Song. 2024. A Uni- fied Temporal Knowledge Graph Reasoning Model Towards Interpolation and Extrapolation.arXiv preprint arXiv:2405.18106(2024)

    work page Pith review arXiv 2024

  6. [6]

    Chanyoung Chung, Jaejun Lee, and Joyce Jiyoung Whang. 2023. Representation learning on hyper-relational and numeric knowledge graphs with transformers. InProceedings of the 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining. 310–322. VITA: Versatile Time Representation Learning for Temporal Hyper-Relational Knowledge Graphs Conference acron...

    work page arXiv 2023

  7. [7]

    Jeffrey Dalton, Laura Dietz, and James Allan. 2014. Entity query feature expansion using knowledge base links. InProceedings of the 37th international ACM SIGIR conference on Research & development in information retrieval. 365–374

    work page 2014

  8. [8]

    Shib Sankar Dasgupta, Swayambhu Nath Ray, and Partha Talukdar. 2018. Hyte: Hyperplane-based temporally aware knowledge graph embedding. InProceedings of the 2018 conference on empirical methods in natural language processing. 2001– 2011

    2018

Show all 68 references
  1. [9]

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

    2018

  2. [10]

    Jacob Devlin. 2018. Bert: Pre-training of deep bidirectional transformers for language understanding.arXiv preprint arXiv:1810.04805(2018)

    2018 arXiv

  3. [11]

    Zifeng Ding, Yunpu Ma, Bailan He, Jingpei Wu, Zhen Han, and Volker Tresp. 2023. A simple but powerful graph encoder for temporal knowledge graph completion. InIntelligent Systems Conference. Springer, 729–747

    2023

  4. [12]

    Zifeng Ding, Jingcheng Wu, Jingpei Wu, Yan Xia, and Volker Tresp. 2023. Ex- ploring Link Prediction over Hyper-Relational Temporal Knowledge Graphs Enhanced with Time-Invariant Relational Knowledge.arXiv preprint arXiv:2307.10219(2023)

    2023 arXiv

  5. [13]

    Bahare Fatemi, Perouz Taslakian, David Vazquez, and David Poole. 2019. Knowledge hypergraphs: Prediction beyond binary relations.arXiv preprint arXiv:1906.00137(2019)

    2019 arXiv

  6. [14]

    Dieter Fensel, Umutcan Şimşek, Kevin Angele, Elwin Huaman, Elias Kärle, Olek- sandra Panasiuk, Ioan Toma, Jürgen Umbrich, Alexander Wahler, Dieter Fensel, et al. 2020. Introduction: what is a knowledge graph?Knowledge graphs: Method- ology, tools and selected use cases(2020), 1–10

    2020

  7. [15]

    Mikhail Galkin, Priyansh Trivedi, Gaurav Maheshwari, Ricardo Usbeck, and Jens Lehmann. 2020. Message passing for hyper-relational knowledge graphs.arXiv preprint arXiv:2009.10847(2020)

    2020 arXiv

  8. [16]

    Alberto Garcia-Duran, Sebastijan Dumančić, and Mathias Niepert. 2018. Learning Sequence Encoders for Temporal Knowledge Graph Completion. InProceedings of the 2018 Conference on Empirical Methods in Natural Language Processing. 4816–4821

    2018

  9. [17]

    Genet Asefa Gesese, Russa Biswas, Mehwish Alam, and Harald Sack. 2021. A survey on knowledge graph embeddings with literals: Which model links better literal-ly?Semantic Web12, 4 (2021), 617–647

    2021

  10. [18]

    Rishab Goel, Seyed Mehran Kazemi, Marcus Brubaker, and Pascal Poupart. 2020. Diachronic embedding for temporal knowledge graph completion. InProceedings of the AAAI conference on artificial intelligence, Vol. 34. 3988–3995

    2020

  11. [19]

    Saiping Guan, Xiaolong Jin, Jiafeng Guo, Yuanzhuo Wang, and Xueqi Cheng

  12. [20]

    Saiping Guan, Xiaolong Jin, Yuanzhuo Wang, and Xueqi Cheng. 2019. Link prediction on n-ary relational data. InThe world wide web conference. 583–593

    2019

  13. [21]

    Lingbing Guo, Zequn Sun, and Wei Hu. 2019. Learning to exploit long-term relational dependencies in knowledge graphs. InInternational conference on machine learning. PMLR, 2505–2514

    2019

  14. [22]

    InProceedings of the 58th annual meeting of the association for computational linguistics

    Neuinfer: Knowledge inference on n-ary facts. InProceedings of the 58th annual meeting of the association for computational linguistics. 6141–6151

  15. [23]

    Guoliang Ji, Shizhu He, Liheng Xu, Kang Liu, and Jun Zhao. 2015. Knowledge graph embedding via dynamic mapping matrix. InProceedings of the 53rd annual meeting of the association for computational linguistics and the 7th international joint conference on natural language proce...

    2015

  16. [24]

    Jaehun Jung, Jinhong Jung, and U Kang. 2020. T-gap: Learning to walk across time for temporal knowledge graph completion.arXiv preprint arXiv:2012.10595 (2020)

    2020 arXiv

  17. [25]

    Zhen Han, Peng Chen, Yunpu Ma, and Volker Tresp. 2020. xerte: Explainable reasoning on temporal knowledge graphs for forecasting future links.arXiv preprint arXiv:2012.15537(2020)

    2020 arXiv

  18. [26]

    Agustinus Kristiadi, Mohammad Asif Khan, Denis Lukovnikov, Jens Lehmann, and Asja Fischer. 2019. Incorporating literals into knowledge graph embeddings. InThe Semantic Web–ISWC 2019: 18th International Semantic Web Conference, Auckland, New Zealand, October 26–30, 2019, Procee...

    2019

  19. [27]

    Agustinus Kristiadi, Mohammad Asif Khan, Denis Lukovnikov, Jens Lehmann, and Asja Fischer. 2019. Incorporating Literals into Knowledge Graph Embeddings. arXiv:1802.00934 [cs.AI] https://arxiv.org/abs/1802.00934

    2019 arXiv

  20. [28]

    Thomas N Kipf and Max Welling. 2016. Semi-supervised classification with graph convolutional networks.arXiv preprint arXiv:1609.02907(2016)

    2016 arXiv

  21. [29]

    Julien Leblay and Melisachew Wudage Chekol. 2018. Deriving validity time in knowledge graph. InCompanion proceedings of the the web conference 2018. 1771–1776

    2018

  22. [30]

    Jiang Li, Xiangdong Su, and Guanglai Gao. 2023. Teast: Temporal knowledge graph embedding via archimedean spiral timeline. InProceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 15460–15474

    2023

  23. [31]

    Timothée Lacroix, Guillaume Obozinski, and Nicolas Usunier. 2020. Ten- sor decompositions for temporal knowledge base completion.arXiv preprint arXiv:2004.04926(2020)

    2020 arXiv

  24. [32]

    Zixuan Li, Xiaolong Jin, Saiping Guan, Wei Li, Jiafeng Guo, Yuanzhuo Wang, and Xueqi Cheng. 2021. Search from history and reason for future: Two-stage reasoning on temporal knowledge graphs.arXiv preprint arXiv:2106.00327(2021)

    2021 arXiv

  25. [33]

    Zixuan Li, Xiaolong Jin, Wei Li, Saiping Guan, Jiafeng Guo, Huawei Shen, Yuanzhuo Wang, and Xueqi Cheng. 2021. Temporal knowledge graph reasoning based on evolutional representation learning. InProceedings of the 44th interna- tional ACM SIGIR conference on research and develo...

    2021

  26. [34]

    Yujia Li, Shiliang Sun, and Jing Zhao. 2022. TiRGN: Time-Guided Recurrent Graph Network with Local-Global Historical Patterns for Temporal Knowledge Graph Reasoning.. InIJCAI. 2152–2158

    2022

  27. [35]

    Lihui Liu, Boxin Du, Jiejun Xu, Yinglong Xia, and Hanghang Tong. 2022. Joint knowledge graph completion and question answering. InProceedings of the 28th ACM SIGKDD conference on knowledge discovery and data mining. 1098–1108

    2022

  28. [36]

    Yu Liu, Quanming Yao, and Yong Li. 2020. Generalizing tensor decomposition for n-ary relational knowledge bases. InProceedings of the web conference 2020. 1104–1114

    2020

  29. [37]

    Yankai Lin, Zhiyuan Liu, Maosong Sun, Yang Liu, and Xuan Zhu. 2015. Learning entity and relation embeddings for knowledge graph completion. InProceedings of the AAAI conference on artificial intelligence, Vol. 29

    2015

  30. [38]

    Johannes Messner, Ralph Abboud, and Ismail Ilkan Ceylan. 2022. Temporal knowledge graph completion using box embeddings. InProceedings of the AAAI Conference on Artificial Intelligence, Vol. 36. 7779–7787

    2022

  31. [39]

    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

  32. [40]

    Yu Liu, Quanming Yao, and Yong Li. 2021. Role-aware modeling for n-ary relational knowledge bases. InProceedings of the Web Conference 2021. 2660– 2671

    2021

  33. [41]

    Jiaxin Pan, Mojtaba Nayyeri, Yinan Li, and Steffen Staab. 2024. HGE: Embedding Temporal Knowledge Graphs in a Product Space of Heterogeneous Geometric Subspaces. InProceedings of the AAAI Conference on Artificial Intelligence, Vol. 38. 8913–8920

    2024

  34. [42]

    Paolo Rosso, Dingqi Yang, and Philippe Cudré-Mauroux. 2020. Beyond triplets: hyper-relational knowledge graph embedding for link prediction. InProceedings of the web conference 2020. 1885–1896

    2020

  35. [43]

    Maximilian Nickel, Volker Tresp, Hans-Peter Kriegel, et al. 2011. A three-way model for collective learning on multi-relational data.. InIcml, Vol. 11. 3104482– 3104584. Conference acronym ’XX, June 03–05, 2018, Woodstock, NY Trovato et al

    2011

  36. [44]

    Harry Shomer, Wei Jin, Juanhui Li, Yao Ma, and Hui Liu. 2023. Learning represen- tations for hyper-relational knowledge graphs. InProceedings of the International Conference on Advances in Social Networks Analysis and Mining. 253–257

    2023

  37. [45]

    Xing Tang, Ling Chen, Hongyu Shi, and Dandan Lyu. 2024. DHyper: A Recurrent Dual Hypergraph Neural Network for Event Prediction in Temporal Knowledge Graphs.ACM Transactions on Information Systems42, 5 (2024), 1–23

    2024

  38. [46]

    Cristian Santini, Genet Asefa Gesese, Silvio Peroni, Aldo Gangemi, Harald Sack, and Mehwish Alam. 2022. A knowledge graph embeddings based approach for author name disambiguation using literals.Scientometrics127, 8 (2022), 4887– 4912

    2022

  39. [47]

    A Vaswani. 2017. Attention is all you need.Advances in Neural Information Processing Systems(2017)

    2017

  40. [48]

    Chenxu Wang, Xin Wang, Zhao Li, Zirui Chen, and Jianxin Li. 2023. Hyconve: A novel embedding model for knowledge hypergraph link prediction with con- volutional neural networks. InProceedings of the ACM Web Conference 2023. 188–198

    2023

  41. [49]

    Théo Trouillon, Johannes Welbl, Sebastian Riedel, Éric Gaussier, and Guillaume Bouchard. 2016. Complex embeddings for simple link prediction. InInternational conference on machine learning. PMLR, 2071–2080

    2016

  42. [50]

    Quan Wang, Haifeng Wang, Yajuan Lyu, and Yong Zhu. 2021. Link prediction on n-ary relational facts: A graph-based approach.arXiv preprint arXiv:2105.08476 (2021)

    2021 arXiv

  43. [51]

    Xiang Wang, Xiangnan He, Yixin Cao, Meng Liu, and Tat-Seng Chua. 2019. Kgat: Knowledge graph attention network for recommendation. InProceedings of the 25th ACM SIGKDD international conference on knowledge discovery & data mining. 950–958

    2019

  44. [52]

    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 engineering29, 12 (2017), 2724–2743

    2017

  45. [53]

    Jianfeng Wen, Jianxin Li, Yongyi Mao, Shini Chen, and Richong Zhang. 2016. On the representation and embedding of knowledge bases beyond binary relations. arXiv preprint arXiv:1604.08642(2016)

    2016 arXiv

  46. [54]

    Jiapeng Wu, Meng Cao, Jackie Chi Kit Cheung, and William L Hamilton. 2020. Temp: Temporal message passing for temporal knowledge graph completion. arXiv preprint arXiv:2010.03526(2020)

    2020 arXiv

  47. [55]

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

    2014

  48. [56]

    Chengjin Xu, Mojtaba Nayyeri, Fouad Alkhoury, Hamed Shariat Yazdi, and Jens Lehmann. 2020. TeRo: A time-aware knowledge graph embedding via temporal rotation.arXiv preprint arXiv:2010.01029(2020)

    2020 arXiv

  49. [57]

    Da Xu, Chuanwei Ruan, Evren Korpeoglu, Sushant Kumar, and Kannan Achan

  50. [58]

    Chengjin Xu, Mojtaba Nayyeri, Fouad Alkhoury, Hamed Shariat Yazdi, and Jens Lehmann. 2019. Temporal knowledge graph embedding model based on additive time series decomposition.arXiv preprint arXiv:1911.07893(2019)

    2019 arXiv

  51. [59]

    Donghan Yu and Yiming Yang. 2021. Improving hyper-relational knowledge graph completion.arXiv preprint arXiv:2104.08167(2021)

    2021 arXiv

  52. [60]

    Richong Zhang, Junpeng Li, Jiajie Mei, and Yongyi Mao. 2018. Scalable instance reconstruction in knowledge bases via relatedness affiliated embedding. InPro- ceedings of the 2018 world wide web conference. 1185–1194

    2018

  53. [61]

    Inductive representation learning on temporal graphs.arXiv preprint arXiv:2002.07962(2020)

    2020 arXiv

  54. [62]

    Bishan Yang, Wen-tau Yih, Xiaodong He, Jianfeng Gao, and Li Deng. 2014. Em- bedding entities and relations for learning and inference in knowledge bases. arXiv preprint arXiv:1412.6575(2014)

    2014 arXiv

  55. [63]

    Cunchao Zhu, Muhao Chen, Changjun Fan, Guangquan Cheng, and Yan Zhang

  56. [65]

    Xuanyu Zhang, Qing Yang, and Dongliang Xu. 2022. TranS: Transition-based knowledge graph embedding with synthetic relation representation.arXiv preprint arXiv:2204.08401(2022)

    2022 arXiv

  57. [66]

    Yuchao Zhang, Xiangjie Kong, Zhehui Shen, Jianxin Li, Qiuhua Yi, Guojiang Shen, and Bo Dong. 2024. A survey on temporal knowledge graph embedding: Models and applications.Knowledge-Based Systems(2024), 112454

    2024

  58. [2020]

    Boxe: A box embedding model for knowledge base completion.Advances in Neural Information Processing Systems33 (2020), 9649–9661

    2020

  59. [2021]

    InProceedings of the AAAI conference on artificial intelligence, Vol

    Learning from history: Modeling temporal knowledge graphs with sequen- tial copy-generation networks. InProceedings of the AAAI conference on artificial intelligence, Vol. 35. 4732–4740. A Details of Baselines and Settings We compare VITA against a sizeable collection of state...

    2018

  60. [2024]

    A Survey on Temporal Knowledge Graph: Representation Learning and Applications.arXiv preprint arXiv:2403.04782(2024)

    2024 arXiv

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.