Pith. sign in

REVIEW 5 major objections 5 minor 82 references

Treating each subject–relation pair as an event chain, GAttNHP learns shared Hawkes excitation through latent groups and forecasts both the next object and its arrival time.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

GAttNHP pairs an attention-based event-chain encoder, a soft-group cross-chain excitation mask, and a non-crossing quantile head to predict entities and occurrence times in temporal knowledge graphs, with large gains on event-heavy datasets but not on all six.

T0 review reviewed 2026-08-02 challenge →

load-bearing objection A useful assembly of known components with strong ICEWS gains, but the abstract overclaims on YAGO and the group mechanism is only defined inside a 16-chain mini-batch, so the core explanation is not yet supported. the 5 major comments →

arxiv 2607.14733 v1 pith:56NMFXWC submitted 2026-07-16 cs.LG stat.ML

GAttNHP: Group Attention Neural Hawkes Process for Extrapolation Reasoning in Temporal Knowledge Graphs

Xiangni Tian , Kaixian Yu , Runpeng Dai , Niansheng Tang , Hongtu Zhu This is my paper
classification cs.LG stat.ML MSC 60G5568T07
keywords temporal knowledge graphextrapolation reasoningHawkes processmutual excitationself-attentionquantile regressionlong-tail eventslink prediction
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

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 a temporal knowledge graph is better modeled as a collection of mutually exciting event chains than as static snapshots, and that cross-chain excitation can be captured cheaply by soft group assignment translated into an attention mask. If right, both what-happens-next and when-it-happens predictions improve substantially, with rare chains gaining the most because they borrow statistical strength from semantically similar chains. Evidence is reported on six benchmarks, with the largest gains on low-frequency chains.

Core claim

GAttNHP claims that a temporal knowledge graph can be recast as marked, mutually exciting event chains, one per subject–relation pair, and that a global group-level Hawkes prior can be converted into an exact attention mask that modulates semantic query–key similarity. Together with a non-crossing quantile head for inter-arrival times, this yields state-of-the-art entity prediction and calibrated time prediction on six TKG benchmarks, with ablations showing the group-excitation module is the primary driver and that its largest gains are on long-tail event chains.

What carries the argument

The central object is the Hawkes-derived analytical attention mask M_{q,k} = log(Φ̄_{u,v}) − γ̄_u Δt, which, when exponentiated inside softmax, becomes a group-Hawkes excitation–decay kernel Φ̄_{u,v} exp(−γ̄_u Δt) that modulates semantic similarity across chains. It is built from a soft group assignment W over G=4 latent groups and globally learned group-to-group excitation matrix Φ and decay vector γ, reducing quadratic cross-chain computation to O(G²).

Load-bearing premise

Cross-chain excitation is estimated from co-occurrences within a random mini-batch of 16 chains, so if truly coupled chains rarely appear together in the same batch, the learned group excitation matrix encodes batch-sampling statistics rather than genuine event dependencies.

What would settle it

Randomly permute chain identities within each mini-batch while preserving each chain's own event sequence, so group assignments become meaningless relative to the actual chains; if the group module's gains on tail chains persist unchanged, the learned excitation is not capturing semantically meaningful cross-chain structure.

Watch this falsifier. Get emailed when new claim-graph text bears on it.

If this is right

  • Entity prediction on ICEWS14 improves to 0.5068 MRR, an absolute gain of 8.25 points over the prior best ECEformer, with Hits@1 nearly 10 points above TLogic.
  • Time prediction MAE is reduced by 47–70% compared with an internal MSE head on ICEWS datasets, while empirical 90% interval coverage stays close to nominal.
  • The group-interaction module's largest gains are on mid- and low-frequency chains: tail-chain MRR rises by 10.7 points over the self-attention-only variant.
  • The method is far lighter than strong baselines, using 2.8–8.5M parameters versus 88–107M and fitting on a single 16 GB GPU.
  • A mean-based time objective makes joint training unstable and collapses entity prediction, whereas the non-crossing quantile head stabilizes optimization.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The group-excitation gains on tail chains may be largely a semantic-smoothing or implicit data-augmentation effect; if so, a simpler group-mean intensity prior could reproduce much of the benefit, a hypothesis the paper does not test.
  • The NCQ head's advantage over MSE may stem as much from replacing quadratically exploding gradients with bounded pinball gradients as from genuinely better distributional calibration; a trimmed-MSE or Huber-loss control would separate these.
  • The framework suggests a general recipe for other sparse event-stream domains: treat each unit as a chain, let a small latent group structure mediate cross-unit excitation, and forecast with a monotone quantile head—potentially applicable to fraud detection, epidemic spread, or social media cascades.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

5 major / 5 minor

Summary. GAttNHP models each (subject, relation) pair in a temporal knowledge graph as a marked event chain and treats TKG extrapolation as a marked temporal point process. The method has three components: (i) a self-attention history encoder (AttNHP) that computes a continuous-time chain embedding; (ii) a group-excitation module that softly assigns chains to G latent groups, learns a group-to-group excitation matrix Φ and decay vector γ, and uses these to form a Hawkes-derived mask in a cross-attention layer over a mini-batch of chains; and (iii) a Non-Crossing Quantile (NCQ) head, adapted from Wu et al. (2023), that predicts monotone quantiles of the inter-arrival distribution via a value/delta layer construction. The model is trained by joint NLL and pinball losses. Experiments report entity-prediction MRR/Hits@k on six datasets and time-prediction MAE/quantile scores on three ICEWS datasets, with ablations and frequency-based analyses.

Significance. The paper is a serious attempt to bring continuous-time mutual-excitation modeling to TKG extrapolation, and it has real strengths: entity baselines are re-run with official code on the same chronological splits; the ablations separate the self-attention, group, and NCQ components; and the efficiency analysis (Table 5) shows that the group mechanism is cheap in parameters and memory. If the claims were fully supported, GAttNHP would be a useful addition to the TKG-forecasting literature and would provide a practical way to share statistical strength across event chains. However, the current manuscript overstates the empirical results, and the central group mechanism is underspecified at inference time; both issues must be resolved before the contribution can be assessed.

major comments (5)
  1. [Abstract/§4.1, Table 6] The abstract and Section 4.1's opening claim ('Across every metric and every dataset, GAttNHP substantially outperforms prior methods') are contradicted by Appendix Table 6. On YAGO, CENET achieves MRR 0.6060 and TLogic 0.5831, both above GAttNHP's 0.5130. Furthermore, time prediction is evaluated only on ICEWS14/18/05-15 (Tables 2-3), so the abstract's 'on six benchmark TKG datasets, GAttNHP improves ... on both entity prediction and time prediction' is not supported for time prediction on GDELT/WIKI/YAGO. Please restrict the claims to the settings actually evaluated and discuss the YAGO gap explicitly in the abstract or conclusion.
  2. [§3.4, Eqs. (15)-(18), (21); §A.2.1; §A.5] Section 3.4 and Eqs. (15)-(18), (21) define the group-excitation term only over a mini-batch of B=16 chains: Eq. (15) forms Φ-bar = W Φ W^T with W in R^{B×G}, Eq. (18) constructs H_shared in R^{B×L×D}, and Eq. (21) applies the Hawkes mask inside a cross-attention softmax over that batch. The test-time batching protocol is not specified anywhere in Section 4 or Appendix A.2.1. Consequently, if test queries are batched arbitrarily, the same query gets different cross-chain contexts (and different ranks) across runs, and with B=1 the group term degenerates to within-chain attention. The ablation Base→Group (Figure 2, Section A.4.1) and the frequency analysis (Table 7, Section A.5) therefore do not isolate 'semantic group excitation' from the generic ability to attend to other chains present in the batch. Please specify the inference-time batching (e.g., fixed global context chains, one batc
  3. [§A.1.1, Table 4] Table 4 in Section A.1.1 reports ICEWS05-15 train/valid/test sizes of 38,692/46,092/46,275. This is inconsistent with the stated chronological 80/10/10 split (validation would be larger than training) and differs from the standard ICEWS05-15 split used in prior work. Please correct the dataset statistics and state whether all reported ICEWS05-15 results use the corrected split; as printed, the main-table ICEWS05-15 numbers cannot be reproduced.
  4. [§3.6, Eqs. (24)-(26)] Section 3.6 presents the Non-Crossing Quantile head as one of the paper's three contributions. The construction—value layer, delta layer, softplus increments, and centered cumulative sum to enforce monotonicity—appears to be the same NCQ module as DNet (Wu et al., 2023), which the text only credits with 'adapted from Wu et al. (2023).' Please state explicitly which elements are new to this paper (e.g., the application to TKG inter-arrival times, the joint loss, or an architectural change). If the head is identical, the contribution claim should be reframed as an adaptation/evaluation and DNet should be cited in the introduction and abstract where the contribution is announced.
  5. [§A.4.1, §A.2.2] Section A.4.1's comparison of NCQ vs. MSE in multi-task training reports a catastrophic entity-prediction collapse for the MSE variant (e.g., ICEWS14 MRR drops from 0.5118 to 0.1124), which is used to argue that NCQ is 'essential.' However, Section A.2.2 reports a hyperparameter grid for β but does not say that the MSE variant's β was re-tuned; a small β that is appropriate for the bounded pinball loss may be very different for the squared-error loss. Please tune β separately for the MSE variant, or report the β values used, so that the comparison is controlled.
minor comments (5)
  1. [§3.7 vs §A.2.1] Section 3.7 says the final model is selected by minimum validation loss, while Section A.2.1 says 'We keep the checkpoint with the lowest training loss.' Please reconcile these statements.
  2. [Abstract] The abstract's 'six benchmark TKG datasets ... on both entity prediction and time prediction' should be revised to reflect that entity prediction is evaluated on six datasets and time prediction on three.
  3. [Appendix preamble] The line 'The appendix is preserved from the original manuscript:' at the start of Appendix A appears to be an editorial note rather than content and should be removed.
  4. [Eq. (23)] The NLL integral in Eq. (23) is approximated with a single-step rectangle rule; please provide the discretization used or a sensitivity check showing that a finer approximation does not change the results.
  5. [Tables 2-3] For the external time-prediction baselines GHT and GHNN, please state explicitly which point prediction (e.g., conditional mean, median, or sampled time) was used to compute MAE, since their native output distributions may differ from the median quantile used for NCQ.

Circularity Check

0 steps flagged

No significant circularity: empirical claims rest on benchmark re-runs and ablations; the reused NCQ head is cited and not load-bearing.

full rationale

The paper's empirical claims are grounded in benchmark re-runs and ablations, not in a self-referential derivation. The self-attention encoder explicitly follows Yang et al. (2022); the group-intensity decomposition explicitly follows GNHP (Fang et al., 2024a); and the NCQ quantile head is explicitly adapted from Wu et al. (2023), with the shared author Hongtu Zhu making it a self-citation. This self-citation is not load-bearing: the time-prediction superiority of NCQ is established in Tables 2–3 by head-to-head comparisons on the same encoder, and the entity-prediction gains are established by Table 1 and the ablations. The identity in Eq. (17), exp(M) = Φ_bar exp(-γ_bar Δt), is a definitional property of the mask, not a derived prediction; the paper presents it as a design rationale, not as an empirical result. The mini-batch implementation (B=16) raises a possible confound about what the group term measures at test time, but this is an experimental limitation, not a circular reduction of a prediction to a fit. No fitted parameter is renamed as a prediction, and no load-bearing claim depends solely on a self-citation. Therefore the derivation chain is self-contained and the circularity score is 0.

Axiom & Free-Parameter Ledger

6 free parameters · 6 axioms · 2 invented entities

The central result rests on fitted quantities and untested assumptions: G=4 groups, the learned excitation matrix Φ and decay γ (called 'priors' but fitted), the assignment MLP, and task weight β are all chosen by search or gradient fitting; the NLL integral in Eq. (23) is a one-point rectangle approximation; cross-chain excitation is limited to 16-chain batches; the time head ignores the queried object (Eq. 6 vs Eq. 24); and the latent groups and Φ have no external falsifiable handle.

free parameters (6)
  • Number of latent groups G = 4
    Grid-searched over {1,2,4,8,16}, selected by validation balance of MRR and MAE (§A.2.2, §A.4.2); the entire group-excitation mechanism depends on this capacity choice.
  • Group-to-group excitation matrix Φ = 4x4, learned
    'Globally learnable' group Hawkes prior (§3.4, Eq. 15); fitted by the joint loss, not a fixed prior.
  • Group decay vector γ = 4-dim, learned
    Same as Φ; sets the exponential decay rate of cross-chain excitation.
  • Time-loss weight β = 0.05 (ICEWS/WIKI/YAGO), 0.001 (GDELT)
    Grid-searched over {1,0.5,0.1,0.05,0.01,0.001}; the joint-training behavior (and the claimed NCQ-vs-MSE difference) depends on this scale.
  • Soft-group assignment MLP + temperature τ = learned
    Eq. (14); maps (e_s, e_r) to group memberships end-to-end; the 'semantics' of groups are whatever this MLP fits.
  • Mini-batch size B = 16
    Grid-searched over {4,8,16,32}; also defines the scope of cross-chain excitation.
axioms (6)
  • ad hoc to paper Single-point rectangle rule adequately approximates the NLL time-integral in Eq. (23).
    Invoked in §3.5 ('We approximate the integral term using the rectangle rule, computed as the product of the time interval and the aggregated intensity at the target time step'); no error bound, no ablation; a biased integral distorts the TPP likelihood that all downstream predictions rest on.
  • domain assumption Cross-chain excitation can be learned from co-occurrence within random mini-batches of B=16 chains.
    §3.4 (Eqs. 15-18) and §A.2.1 (batch size 16); the model never computes excitation across the full chain population.
  • domain assumption G=4 latent groups with MLP assignment from static (e_s, e_r) embeddings capture the semantically meaningful groupings.
    Eq. (14); G chosen by validation over {1,2,4,8,16}; grouping quality is only indirectly validated through downstream metrics.
  • domain assumption TKG extrapolation benchmark protocol (chronological 80/10/10 splits; time-aware raw MRR; re-run baselines on the same splits) supports the comparisons.
    §4.1 and §A.2.1-3; raw MRR is unusual versus the filtered MRR common in the TKG literature, and baseline re-runs are not verifiable from the paper alone.
  • ad hoc to paper The NCQ head's quantiles target the next event on the chain, not the next occurrence of the queried object.
    Eq. (6) defines the task conditioned on (s,r,o) but Eq. (24) conditions only on chain-level z_u(t); the paper never states which protocol is used in Tables 2-3, yet the model literally cannot condition on o.
  • standard math Standard multivariate Hawkes and neural point-process formalism, softplus positivity, and stochastic gradient training.
    Background material (Eqs. 1-3); unproblematic.
invented entities (2)
  • Latent semantic groups (G=4 soft clusters over (s,r) chains) no independent evidence
    purpose: Let chains share excitation patterns through group memberships instead of pairwise computation (challenge ii)
    No external falsifiable handle; grouping quality is validated only by downstream benchmark gains and an internal case study (§A.5.1).
  • Group-level Hawkes excitation matrix Φ and decay vector γ ('globally learnable priors') no independent evidence
    purpose: Macro-level cross-chain mutual excitation prior folded into the attention mask (Eq. 16)
    Fitted parameters, not priors; no external calibration or prediction that could falsify them.

reviewed 2026-08-02 · how reviews work

0 comments
Cite this review

Pith. "Pith review of GAttNHP: Group Attention Neural Hawkes Process for Extrapolation Reasoning in Temporal Knowledge Graphs." pith.science (2026). https://pith.science/paper/56NMFXWC

@misc{pith2026260714733,
  author       = {Pith},
  title        = {Pith review of: GAttNHP: Group Attention Neural Hawkes Process for Extrapolation Reasoning in Temporal Knowledge Graphs},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/56NMFXWC}},
  note         = {Machine review of arXiv:2607.14733}
}
Share X Bluesky LinkedIn Reddit HN
read the original abstract

Temporal Knowledge Graphs (TKGs) record how facts evolve over time, but forecasting future events on a TKG remains difficult for three reasons: (i) long-range temporal dependencies are hard to encode; (ii) events on different chains mutually excite or inhibit one another in ways that snapshot-level models cannot express; and (iii) inter-arrival times are heavy-tailed and statistically sparse, so deterministic time predictors are unreliable. We address these three issues with a single framework, the \textbf{Group Attention Neural Hawkes Process (GAttNHP)}, built around three matched components. First, a self-attention encoder casts each subject--relation chain as a continuous-time point process and captures the lingering excitation of distant history. Second, a semantic soft-grouping module turns globally learnable Hawkes priors into an analytical cross-attention mask, so chains share excitation patterns through their latent group memberships rather than through exhaustive pairwise computation. Third, a Non-Crossing Quantile (NCQ) regression head replaces mean-based time prediction, providing calibrated, monotonically ordered quantile estimates that remain stable under heavy-tailed inter-arrival distributions. On six benchmark TKG datasets, GAttNHP improves over state-of-the-art baselines on both entity prediction and time prediction, and ablations confirm that its largest gains arise on the long-tail event chains where existing models fail most severely.

Figures

Figures reproduced from arXiv: 2607.14733 by Hongtu Zhu, Kaixian Yu, Niansheng Tang, Runpeng Dai, Xiangni Tian.

Figure 1
Figure 1. Figure 1: From a temporal knowledge graph to a marked point process. (a) Conventionally, a TKG is [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Ablation study results across ICEWS14, ICEWS18, and ICEWS05-15 datasets. We [PITH_FULL_IMAGE:figures/full_fig_p015_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Sensitivity analysis of the latent group number ( [PITH_FULL_IMAGE:figures/full_fig_p017_3.png] view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

82 extracted references · 3 linked inside Pith

  1. [1]

    1994 , edition =

    LaTeX: A Document Preparation System , author =. 1994 , edition =

    1994

  2. [2]

    2016 , publisher =

    Deep Learning , author =. 2016 , publisher =

    2016

  3. [3]

    arXiv:2211.10738 , year =

    Liang, Ke and Liu, Yue and Zhou, Sihang and Liu, Xinwang and Tu, Wenxuan , title =. arXiv:2211.10738 , year =

    Pith/arXiv arXiv

  4. [4]

    A survey on knowledge graphs: Representation, acquisition, and applications , year =

    Ji, Shaoxiong and Pan, Shirui and Cambria, Erik and Marttinen, Pekka and Philip, S Yu , journal =. A survey on knowledge graphs: Representation, acquisition, and applications , year =

  5. [5]

    2016 , isbn =

    Zhang, Fuzheng and Yuan, Nicholas Jing and Lian, Defu and Xie, Xing and Ma, Wei-Ying , title =. 2016 , isbn =

    2016

  6. [6]

    and Rangwala, H

    Deng, S. and Rangwala, H. and Ning, Y. , title =. Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining , year =

  7. [7]

    ACM Transactions on Information Systems (TOIS) , volume=

    Temporal relational ranking for stock prediction , author=. ACM Transactions on Information Systems (TOIS) , volume=

  8. [8]

    TEQUILA: Temporal Question Answering over Knowledge Bases , year =

    Jia, Zhen and Abujabal, Abdalghani and Saha Roy, Rishiraj and Str\". TEQUILA: Temporal Question Answering over Knowledge Bases , year =

  9. [9]

    Learning Sequence Encoders for Temporal Knowledge Graph Completion , year =

    Garcia-Duran, Alberto and Dumancic, Sebastijan and Niepert, Mathias , booktitle =. Learning Sequence Encoders for Temporal Knowledge Graph Completion , year =

  10. [10]

    Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing , year =

    Garca-Duran, Alberto and Dumancic, Sebastijan and Niepert, Mathias , title =. Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing , year =

    2018

  11. [11]

    Tensor Decompositions for Temporal Knowledge Base Completion , year =

    Lacroix, Timothee and Obozinski, Guillaume and Usunier, Nicolas , booktitle =. Tensor Decompositions for Temporal Knowledge Base Completion , year =

  12. [12]

    ChronoR: rotation based temporal knowledge graph embedding , year =

    Sadeghian, Ali and Armandpour, Mohammadreza and Colas, Anthony and Wang, Daisy Zhe , journal=. ChronoR: rotation based temporal knowledge graph embedding , year =

  13. [13]

    Tucker decomposition-based temporal knowledge graph completion , volume =

    Shao, Pengpeng and Zhang, Dawei and Yang, Guohua and Tao, Jianhua and Che, Feihu and Liu, Tong , journal =. Tucker decomposition-based temporal knowledge graph completion , volume =

  14. [14]

    Liao, Siyuan and Liang, Shangsong and Meng, Zaiqiao and Zhang, Qiang , title =

  15. [15]

    Hyte: Hyperplane-based temporally aware knowledge graph embedding , publisher =

    Dasgupta, Shib Sankar and Ray, Swayambhu Nath and Talukdar, Partha , booktitle =. Hyte: Hyperplane-based temporally aware knowledge graph embedding , publisher =

  16. [16]

    Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)

    Chen, Kai and Wang, Ye and Li, Yitong and Li, Aiping , title =. Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)

  17. [17]

    The Semantic Web -- ISWC 2020

    Xu, Chenjin and Nayyeri, Mojtaba and Alkhoury, Fouad and Yazdi, Hamed and Lehmann, Jens , title =. The Semantic Web -- ISWC 2020

    2020

  18. [18]

    Diachronic embedding for temporal knowledge graph completion , volume=

    Goel, Rishab and Kazemi, Seyed Mehran and Brubaker, Marcus and Poupart, Pascal , journal=. Diachronic embedding for temporal knowledge graph completion , volume=

  19. [19]

    Know-evolve: Deep temporal reasoning for dynamic knowledge graphs , volume =

    Trivedi, Rakshit and Dai, Hanjun and Wang, Yichen and Song, Le , booktitle =. Know-evolve: Deep temporal reasoning for dynamic knowledge graphs , volume =

  20. [20]

    Evolvegcn: Evolving graph convolutional networks for dynamic graphs , year =

    Pareja, Aldo and Domeniconi, Giacomo and Chen, Jie and Ma, Tengfei and Suzumura, Toyotaro and Kanezashi, Hiroki and Kaler, Tim and Schardl, Tao and Leiserson, Charles , booktitle =. Evolvegcn: Evolving graph convolutional networks for dynamic graphs , year =

  21. [21]

    Temporal knowledge graph reasoning based on evolutional representation learning , year =

    Li, Zixuan and Jin, Xiaolong and Li, Wei and Guan, Saiping and Guo, Jiafeng and Shen, Huawei and Wang, Yuanzhuo and Cheng, Xueqi , booktitle =. Temporal knowledge graph reasoning based on evolutional representation learning , year =

  22. [22]

    Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP)

    Jin, Woojeong and Qu, Meng and Jin, Xisen and Ren, Xiang , booktitle = "Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP)", address = "Online", pages = "6669--6683", publisher = "Association for Computational Linguistics", title =

    2020

  23. [23]

    Learning from history: Modeling temporal knowledge graphs with sequential copy-generation networks , year =

    Zhu, Cunchao and Chen, Muhao and Fan, Changjun and Cheng, Guangquan and Zhang, Yan , booktitle =. Learning from history: Modeling temporal knowledge graphs with sequential copy-generation networks , year =

  24. [24]

    Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)

    Li, Zixuan and Guan, Saiping and Jin, Xiaolong and Peng, Weihua and Lyu, Yajuan and Zhu, Yong and Bai, Long and Li, Wei and Guo, Jiafeng and Cheng, Xueqi , booktitle = "Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)", pages = "290--296", address = "Dublin, Ireland", publisher = "Association...

  25. [25]

    Temporal Knowledge Graph Reasoning with Historical Contrastive Learning , volume=

    Xu, Yi and Ou, Junjie and Xu, Hui and Fu, Luoyi , booktitle =. Temporal Knowledge Graph Reasoning with Historical Contrastive Learning , volume=

  26. [26]

    , pages =

    Li, Zixuan and Jin, Xiaolong and Guan, Saiping and Li, Wei and Guo, Jiafeng and Wang, Yuanzhuo and Cheng, Xueqi , booktitle = "Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers)", pages = "4732--4743", address = "Online",...

  27. [27]

    Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP)

    Lv, Xin and Han, Xu and Hou, Lei and Li, Juanzi and Liu, Zhiyuan and Zhang, Wei and Zhang, Yichi and Kong, Hao and Wu, Suhui , booktitle = "Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP)", title =

    2020

  28. [28]

    Multi-hop reasoning over paths in temporal knowledge graphs using reinforcement learning , year =

    Bai, Luyi and Yu, Wenting and Chen, Mingzhuo and Ma, Xiangnan , journal =. Multi-hop reasoning over paths in temporal knowledge graphs using reinforcement learning , year =

  29. [29]

    Tlogic: Temporal logical rules for explainable link forecasting on temporal knowledge graphs , volume =

    Liu, Yushan and Ma, Yunpu and Hildebrandt, Marcel and Joblin, Mitchell and Tresp, Volker , booktitle =. Tlogic: Temporal logical rules for explainable link forecasting on temporal knowledge graphs , volume =

  30. [30]

    Graph H awkes Transformer for Extrapolated Reasoning on Temporal Knowledge Graphs

    Sun, Haohai and Geng, Shangyi and Zhong, Jialun and Hu, Han and He, Kun. Graph H awkes Transformer for Extrapolated Reasoning on Temporal Knowledge Graphs. Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing. 2022

    2022

  31. [31]

    Findings of the Association for Computational Linguistics: ACL 2023 , year=

    Learning joint structural and temporal contextualized knowledge embeddings for temporal knowledge graph completion , author=. Findings of the Association for Computational Linguistics: ACL 2023 , year=

    2023

  32. [32]

    Proceedings of the 47th International ACM SIGIR Conference on Research and Development in Information Retrieval , pages=

    Transformer-based reasoning for learning evolutionary chain of events on temporal knowledge graph , author=. Proceedings of the 47th International ACM SIGIR Conference on Research and Development in Information Retrieval , pages=

  33. [33]

    Biometrika , volume=

    Spectra of some self-exciting and mutually exciting point processes , author=. Biometrika , volume=. 1971 , publisher=

    1971

  34. [34]

    Proceedings of the Thirtieth International Joint Conference on Artificial Intelligence , pages =

    Neural Temporal Point Processes: A Review , author =. Proceedings of the Thirtieth International Joint Conference on Artificial Intelligence , pages =

  35. [35]

    Stochastic Processes and Their Applications , volume=

    A self-correcting point process , author=. Stochastic Processes and Their Applications , volume=. 1979 , publisher=

    1979

  36. [37]

    Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining , pages=

    Recurrent marked temporal point processes: Embedding event history to vector , author=. Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining , pages=

  37. [38]

    Proceedings of the 31st International Conference on Neural Information Processing Systems , pages=

    The neural hawkes process: A neurally self-modulating multivariate point process , author =. Proceedings of the 31st International Conference on Neural Information Processing Systems , pages=

  38. [39]

    Proceedings of the 37th International Conference on Machine Learning , pages =

    Transformer Hawkes Process , author =. Proceedings of the 37th International Conference on Machine Learning , pages =

  39. [40]

    Proceedings of the International Conference on Learning Representations , year =

    Transformer Embeddings of Irregularly Spaced Events and Their Participants , author =. Proceedings of the International Conference on Learning Representations , year =

  40. [41]

    Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP)

    Han, Zhen and Chen, Peng and Ma, Yunpu and Tresp, Volker , booktitle = "Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP)", title =

    2020

  41. [42]

    Journal of the American Statistical Association , volume =

    Group Network Hawkes Process , author =. Journal of the American Statistical Association , volume =

  42. [43]

    Proceedings of the 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining , pages=

    DNet: distributional network for distributional individualized treatment effects , author=. Proceedings of the 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining , pages=

  43. [44]

    Translating embeddings for modeling multi-relational data , publisher =

    Bordes, Antoine and Usunier, Nicolas and Garcia-Duran, Alberto and Weston, Jason and Yakhnenko, Oksana , journal =. Translating embeddings for modeling multi-relational data , publisher =

  44. [45]

    Embedding entities and relations for learning and inference in knowledge bases , year =

    Yang, Bishan and Yih, Wen-tau and He, Xiaodong and Gao, Jianfeng and Deng, Li , journal =. Embedding entities and relations for learning and inference in knowledge bases , year =

  45. [46]

    Complex embeddings for simple link prediction , year =

    Trouillon, Th. Complex embeddings for simple link prediction , year =. Proceedings of the 33rd International Conference on International Conference on Machine Learning , volume =

  46. [47]

    Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers) , pages=

    Htccn: Temporal causal convolutional networks with hawkes process for extrapolation reasoning in temporal knowledge graphs , author=. Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers) , pages=

    2024

  47. [48]

    IEEE Transactions on Knowledge and Data Engineering , volume=

    THCN: A Hawkes process based temporal causal convolutional network for extrapolation reasoning in temporal knowledge graphs , author=. IEEE Transactions on Knowledge and Data Engineering , volume=. 2024 , publisher=

    2024

  48. [49]

    AKBC , year=

    Graph Hawkes Neural Network for Forecasting on Temporal Knowledge Graphs , author=. AKBC , year=

  49. [50]

    International Conference on Artificial Intelligence and Statistics (AISTATS) , year=

    Learning Quantile Functions for Temporal Point Processes with Recurrent Neural Splines , author=. International Conference on Artificial Intelligence and Statistics (AISTATS) , year=

  50. [51]

    Translating embeddings for modeling multi-relational data

    Antoine Bordes, Nicolas Usunier, Alberto Garcia-Duran, Jason Weston, and Oksana Yakhnenko. Translating embeddings for modeling multi-relational data. In Proc.\ of NeurIPS, volume 2, pages 2787--2795, 2013

    2013

  51. [52]

    THCN : A H awkes process based temporal causal convolutional network for extrapolation reasoning in temporal knowledge graphs

    Tingxuan Chen, Jun Long, Zidong Wang, Shuai Luo, Jincai Huang, and Liu Yang. THCN : A H awkes process based temporal causal convolutional network for extrapolation reasoning in temporal knowledge graphs. IEEE Trans.\ Knowl.\ Data Eng., 36(12):9374--9387, 2024a

  52. [53]

    HTCCN : Temporal causal convolutional networks with H awkes process for extrapolation reasoning in temporal knowledge graphs

    Tingxuan Chen, Jun Long, Liu Yang, Zidong Wang, Yongheng Wang, and Xiongnan Jin. HTCCN : Temporal causal convolutional networks with H awkes process for extrapolation reasoning in temporal knowledge graphs. In Proc.\ NAACL-HLT, pages 4056--4066, 2024b

  53. [54]

    Recurrent marked temporal point processes: Embedding event history to vector

    Nan Du, Hanjun Dai, Rakshit Trivedi, Utkarsh Upadhyay, Manuel Gomez-Rodriguez, and Le Song. Recurrent marked temporal point processes: Embedding event history to vector. In Proc.\ of KDD, pages 1555--1564, 2016

    2016

  54. [55]

    Group network H awkes process

    Guanhua Fang, Ganggang Xu, Haochen Xu, Xuening Zhu, and Yongtao Guan. Group network H awkes process. Journal of the American Statistical Association, 119(547):2328--2344, 2024a

  55. [56]

    Transformer-based reasoning for learning evolutionary chain of events on temporal knowledge graph

    Zhiyu Fang, Shuai-Long Lei, Xiaobin Zhu, Chun Yang, Shi-Xue Zhang, Xu-Cheng Yin, and Jingyan Qin. Transformer-based reasoning for learning evolutionary chain of events on temporal knowledge graph. In Proc.\ of SIGIR, pages 70--79, 2024b

  56. [57]

    Learning joint structural and temporal contextualized knowledge embeddings for temporal knowledge graph completion

    Yifu Gao, Yongquan He, Zhigang Kan, Yi Han, Linbo Qiao, and Dongsheng Li. Learning joint structural and temporal contextualized knowledge embeddings for temporal knowledge graph completion. In Findings of ACL, 2023

    2023

  57. [58]

    Learning sequence encoders for temporal knowledge graph completion

    Alberto Garcia-Duran, Sebastijan Dumancic, and Mathias Niepert. Learning sequence encoders for temporal knowledge graph completion. In Proc.\ of EMNLP, pages 4816--4821, 2018

    2018

  58. [59]

    Diachronic embedding for temporal knowledge graph completion

    Rishab Goel, Seyed Mehran Kazemi, Marcus Brubaker, and Pascal Poupart. Diachronic embedding for temporal knowledge graph completion. In Proc.\ of AAAI, volume 34, 2020

    2020

  59. [60]

    Alan G. Hawkes. Spectra of some self-exciting and mutually exciting point processes. Biometrika, 58(1):83--90, 1971

    1971

  60. [61]

    A self-correcting point process

    Valerie Isham and Mark Westcott. A self-correcting point process. Stochastic Process.\ Appl., 8(3):335--347, 1979

    1979

  61. [62]

    Recurrent event network: Autoregressive structure inference over temporal knowledge graphs

    Woojeong Jin, Meng Qu, Xisen Jin, and Xiang Ren. Recurrent event network: Autoregressive structure inference over temporal knowledge graphs. In Proc.\ of EMNLP, pages 6669--6683, 2020

    2020

  62. [63]

    Tensor decompositions for temporal knowledge base completion

    Timoth\'ee Lacroix, Guillaume Obozinski, and Nicolas Usunier. Tensor decompositions for temporal knowledge base completion. In Proc.\ of ICLR, 2020

    2020

  63. [64]

    Search from history and reason for future: Two-stage reasoning on temporal knowledge graphs

    Zixuan Li, Xiaolong Jin, Saiping Guan, Wei Li, Jiafeng Guo, Yuanzhuo Wang, and Xueqi Cheng. Search from history and reason for future: Two-stage reasoning on temporal knowledge graphs. In Proc.\ of ACL-IJCNLP, pages 4732--4743, 2021a

  64. [65]

    Temporal knowledge graph reasoning based on evolutional representation learning

    Zixuan Li, Xiaolong Jin, Wei Li, Saiping Guan, Jiafeng Guo, Huawei Shen, Yuanzhuo Wang, and Xueqi Cheng. Temporal knowledge graph reasoning based on evolutional representation learning. In Proc.\ of SIGIR, pages 408--417, 2021b

  65. [66]

    Complex evolutional pattern learning for temporal knowledge graph reasoning

    Zixuan Li, Saiping Guan, Xiaolong Jin, Weihua Peng, Yajuan Lyu, Yong Zhu, Long Bai, Wei Li, Jiafeng Guo, and Xueqi Cheng. Complex evolutional pattern learning for temporal knowledge graph reasoning. In Proc.\ of ACL, pages 290--296, 2022

    2022

  66. [67]

    TLogic : Temporal logical rules for explainable link forecasting on temporal knowledge graphs

    Yushan Liu, Yunpu Ma, Marcel Hildebrandt, Mitchell Joblin, and Volker Tresp. TLogic : Temporal logical rules for explainable link forecasting on temporal knowledge graphs. In Proc.\ of AAAI, volume 36, pages 4120--4127, 2022

    2022

  67. [68]

    The neural H awkes process: A neurally self-modulating multivariate point process

    Hongyuan Mei and Jason Eisner. The neural H awkes process: A neurally self-modulating multivariate point process. In Proc.\ of NeurIPS, pages 6754--6764, 2017

    2017

  68. [69]

    ChronoR : Rotation based temporal knowledge graph embedding

    Ali Sadeghian, Mohammadreza Armandpour, Anthony Colas, and Daisy Zhe Wang. ChronoR : Rotation based temporal knowledge graph embedding. In Proc.\ of AAAI, volume 35, pages 6471--6479, 2021

    2021

  69. [70]

    Neural temporal point processes: A review

    Oleksandr Shchur, Ali Caner Turkmen, Tim Januschowski, and Stephan G\"unnemann. Neural temporal point processes: A review. In Proc.\ of IJCAI, pages 4585--4593, 2021

    2021

  70. [71]

    Graph H awkes transformer for extrapolated reasoning on temporal knowledge graphs

    Haohai Sun, Shangyi Geng, Jialun Zhong, Han Hu, and Kun He. Graph H awkes transformer for extrapolated reasoning on temporal knowledge graphs. In Proc.\ of EMNLP, pages 7481--7493, 2022

    2022

  71. [72]

    Know-evolve: Deep temporal reasoning for dynamic knowledge graphs

    Rakshit Trivedi, Hanjun Dai, Yichen Wang, and Le Song. Know-evolve: Deep temporal reasoning for dynamic knowledge graphs. In Proc.\ of ICML, volume 70, pages 3462--3471, 2017

    2017

  72. [73]

    Complex embeddings for simple link prediction

    Th\'eo Trouillon, Johannes Welbl, Sebastian Riedel, \'Eric Gaussier, and Guillaume Bouchard. Complex embeddings for simple link prediction. In Proc.\ of ICML, volume 48, pages 2071--2080, 2016

    2071

  73. [74]

    DNet : Distributional network for distributional individualized treatment effects

    Guojun Wu, Ge Song, Xiaoxiang Lv, Shikai Luo, Chengchun Shi, and Hongtu Zhu. DNet : Distributional network for distributional individualized treatment effects. In Proc.\ of KDD, pages 5215--5224, 2023

    2023

  74. [75]

    Temporal knowledge graph completion based on time series Gaussian embedding

    Chenjin Xu, Mojtaba Nayyeri, Fouad Alkhoury, Hamed Yazdi, and Jens Lehmann. Temporal knowledge graph completion based on time series Gaussian embedding. In ISWC, pages 654--671, 2020

    2020

  75. [76]

    Temporal knowledge graph reasoning with historical contrastive learning

    Yi Xu, Junjie Ou, Hui Xu, and Luoyi Fu. Temporal knowledge graph reasoning with historical contrastive learning. In Proc.\ of AAAI, volume 37, pages 4765--4773, 2023

    2023

  76. [77]

    Embedding entities and relations for learning and inference in knowledge bases

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

    Pith/arXiv arXiv 2014

  77. [78]

    Transformer embeddings of irregularly spaced events and their participants

    Chen Yang, Hongyuan Mei, and Jason Eisner. Transformer embeddings of irregularly spaced events and their participants. In Proc.\ of ICLR, 2022

    2022

  78. [79]

    Self-attentive H awkes processes

    Qiang Zhang, Aldo Lipani, Omer Kirnap, and Emine Yilmaz. Self-attentive H awkes processes. arXiv:1907.07561, 2019

    Pith/arXiv arXiv 1907

  79. [80]

    Learning from history: Modeling temporal knowledge graphs with sequential copy-generation networks

    Cunchao Zhu, Muhao Chen, Changjun Fan, Guangquan Cheng, and Yan Zhang. Learning from history: Modeling temporal knowledge graphs with sequential copy-generation networks. In Proc.\ of AAAI, volume 35, pages 4732--4740, 2021

    2021

  80. [81]

    Transformer H awkes process

    Simiao Zuo, Hao Jiang, Zichong Li, Tuo Zhao, and Hongyuan Zha. Transformer H awkes process. In Proc.\ of ICML, volume 119, pages 11692--11702, 2020

    2020

Showing first 80 references.

This paper was first reviewed by deepseek-v4-flash on August 2, 2026.