Pith. sign in

REVIEW 4 major objections 5 minor 1 cited by

Towards Anomaly Detection on Relational Data

T0 review · 4 major / 5 minor · reviewed 2026-08-04 · deepseek-v4-flash

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

Pith's one-line read RelAD detects anomalies in relational databases by reconstructing both attribute values and the typed links between tables, and it outperforms existing methods on six benchmark datasets.

desk verdict New problem formulation with a sensible method, but the under-specified hyperparameter tuning makes the headline AUROC numbers questionable. read the letter →

arxiv 2606.18621 v2 pith:MHLZJ7PI submitted 2026-06-17 cs.LG

Shiyuan Li , Yunfeng Zhao , Yue Tan , Qingfeng Chen , Yixin Liu , Shirui Pan This is my paper · ORCID
classification cs.LG
keywords relationalanomalydetectiondatabasesreconstruction-basedmulti-relationaledgereconstructionconditionalsparsegatinginjectionbenchmarkunsupervisedlearningheterogeneousforeign-keyrelations
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 argues that anomaly detection on relational databases deserves its own treatment, because flattening tables into feature vectors destroys foreign-key semantics and homogeneous-graph methods blur relation types. It proposes RelAD, an unsupervised reconstruction framework that scores each entity by how badly its attributes and its typed connections can be rebuilt from learned normal patterns. A conditional sparse gate suppresses redundant multi-table attributes, while a dual-view edge-reconstruction branch checks whether an entity's connections are consistent with both its own profile and its behavioral context. The paper constructs six relational benchmarks with injected fraud-like anomalies and reports that RelAD beats all tabular and graph baselines on every dataset. If the results hold, relational fraud detection can move away from hand-built features and treat the database schema itself as the signal.

What carries the argument

The central machinery is the pair of reconstruction modules. Conditional sparse-gated attribute reconstruction generates a mask per semantic block, gates the input, encodes it, decodes each block separately, and scores anomalies by the average of top-K normalized block residuals. Dual-view multi-relational edge reconstruction encodes the target entity from its self profile and its child-table behavioral profile independently, encodes neighbor entities per relation type, and reconstructs observed edges via a dot-product softplus loss; scores are top-K normalized average positive-edge negative log-likelihoods per branch and view. These are combined by a hierarchical fusion with hyperparameters

What would settle it

Run RelAD against the same baselines on a relational dataset with ground-truth fraud labels from operational logs rather than injected anomalies, keeping the schema and evaluation protocol identical; if RelAD no longer beats the best baseline, the injection-based evidence is not conclusive. A cheaper check is to construct a benchmark whose anomalies are global attribute shifts or added (not replaced) edges, which deliberately violate the injection design principles, and compare performance.

Watch

Extended reading notes

Core claim

RelAD's central claim is that a normal entity in a relational database can explain (reconstruct) its own attributes and its typed edges, while an anomalous entity cannot. Attribute reconstruction operates on semantic blocks: child-table aggregates and central-table fields get separate conditional masks, separate decoders, and top-K normalized block residuals so sparse local deviations are not drowned out by global reconstruction error. Edge reconstruction treats each foreign-key relation separately and from two views — the central-table self profile and the aggregated child-table behavioral profile — using a shared relation-specific neighbor encoder and a softplus negative-log-likelihood los

Load-bearing premise

The load-bearing premise is that the dataset-specific anomaly injections mimic how anomalies actually appear in real relational data; if real fraud looks different, the benchmark comparisons do not establish real-world transfer.

Editorial extensions

If this is right

  • On all six benchmark datasets, RelAD attains the best AUROC and AUPRC among nine tabular and graph baselines, with relative AUROC gains over the strongest baseline exceeding 10% on four datasets.
  • Ablations show that removing multi-relational edge reconstruction causes the largest performance drop on Amazon, ArXiv, and HM, indicating that typed connection reconstruction is the most critical component.
  • The dual-view design matters: removing the child-profile view hurts more on HM, while removing the self-profile view hurts more on ArXiv; neither view alone is sufficient.
  • Conditional gating and block-specific decoding both improve results, confirming that suppressing redundant cross-table attributes and preserving localized semantic deviations helps detection.
  • Inference cost scales linearly with the total number of target entities, neighbor entities, and edges, so the method remains practical on large relational databases.

Reading between the lines

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

  • The benchmark injections focus on rare local attribute swaps and redirected edges; if real-world anomalies are broad, gradual changes, the reported margin over baselines may not transfer to operational settings.
  • Because the framework ignores temporal dynamics — a limitation the paper itself flags — extending RelAD to dynamic relational databases with evolving entities and edges is a natural next step.
  • The same dual-view reconstruction principle could be reused for a relational anomaly detection foundation model: pretrain on diverse schemas, then score unseen tables at test time without retraining.
  • A testable extension would build anomaly suites that separate attribute-only, edge-only, and mixed anomalies, then check whether the fusion weights alpha and beta need dataset-specific calibration to maintain performance.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 5 minor

Summary. The paper formalizes relational anomaly detection (RAD), the task of scoring entities in a central table of a relational database using both multi-table attributes and primary/foreign-key relationships. The proposed method, RelAD, combines (i) a conditional sparse-gated attribute reconstruction module that selectively gates feature blocks and computes top-K block-level residuals, and (ii) a dual-view multi-relational edge reconstruction module that scores edges through both self-profile and child-profile encoders. The two signals are fused via two scalar weights. The authors construct six benchmark datasets by injecting anomalies into RelBench/RelBench v2 databases, and report AUROC/AUPRC comparisons against tabular and homogeneous-graph anomaly detection baselines. The central claim is that RelAD consistently outperforms all baselines while remaining efficient.

Significance. If the reported results are valid, the paper makes a useful contribution: it is among the first to address anomaly detection directly on relational databases, and it provides a concrete framework, code, and a benchmark suite with systematic anomaly injection. The method is well specified, with complete equations, algorithms, complexity analysis, and mean±std results in the appendix. The design of block-aware conditional gating and dual-view edge reconstruction is reasonable and, in the ablation study, each component appears to contribute. However, the load-bearing evidence for consistent superiority is currently weakened by two evaluation-protocol issues: hyperparameter selection may use test labels, and no relational/h heterogeneous-graph baseline is compared. The benchmark validity also rests entirely on the authors' own injection rules, which align closely with RelAD's inductive biases. These concerns are fixable within the scope of a revision.

major comments (4)
  1. [Sec. 4.1 and Appendix E] The paper states that hyperparameters (learning rate, weight decay, batch size, λ_s, α, β) are tuned by random search per dataset, but does not mention any validation split. Since the task is unsupervised and anomaly labels are used only for evaluation, tuning on the test labels would leak label information and directly inflate the reported AUROC/AUPRC. This threatens the central claim of consistent outperformance in Table 1. Please clarify how hyperparameters were selected, and ideally report results for a fixed validation protocol that does not use test labels. The same protocol must be applied to all baselines.
  2. [Sec. 4.1, Table 1] No heterogeneous-graph or relational deep learning baseline is compared. The paper's argument is that RelAD is better than flattened-tabular and homogeneous-graph approaches, but the method's core novelty is multi-relational modeling. Without comparing to a relational baseline such as R-GCN, RelGNN, or an adapted relational deep learning model, the necessity of the dual-view multi-relational edge-reconstruction design is not established. The ablation 'w/o Relation' shows that removing this component hurts, but it does not show that RelAD's specific relational encoder is better than alternative relational architectures.
  3. [Appendix D and Tables 1/4/5] All evaluation is on self-injected anomalies. The injection rules are designed by the authors and directly encode the kinds of local attribute deviations and relation-specific edge redirections that RelAD is designed to detect. For example, the Amazon injection replaces review categories based on user history, and the ArXiv injection redirects citation edges to beacon papers. This makes the benchmark a test of whether RelAD matches its author-designed injection schemes, not necessarily whether it transfers to real relational fraud. The reported margins may not hold under other anomaly-generating processes. I recommend adding at least one independent injection protocol or a real-world labeled relational dataset, and reporting results under multiple injection configurations.
  4. [Eq. (17), Eq. (7)] The sparsity penalty in Eq. (7) is an L1-style penalty on gate values, not a constraint that forces a specific number of active gates. The paper repeatedly describes the gating as selecting a 'compact subset' of attributes, but the actual objective only encourages small gate values. This is a technical mismatch between the description and the implementation. Please clarify whether the learned gates are actually sparse after training, e.g., by reporting the fraction of gates below a threshold, or adjust the wording to describe the mechanism as soft shrinkage rather than selection.
minor comments (5)
  1. [Sec. 4.1 and Table 3] The text says the anomaly ratio is set to 5% across all datasets, but Table 3 reports Avito at 2.77%. If this is due to the validity constraints in Appendix D, please state this explicitly in the main text.
  2. [Sec. 4.2, 'Hyperparameter Analysis'] The sentence 'The reliances on the two weights differ notably' contains a typo; 'reliances' should be 'reliance'.
  3. [Figures 3 and 5] The heatmaps use different color scales across panels, which makes cross-dataset comparison difficult. It would help to use a shared color scale or annotate the color range on each panel.
  4. [Sec. 3.2, Eq. (15)] Uniform negative sampling over all neighbor entities is used without discussion. Since each relation type has different degree distributions, sampling negatives uniformly may not provide the most informative contrast. A brief discussion or an ablation on negative-sampling strategy would strengthen the paper.
  5. [General] The paper claims that RelAD 'consistently outperforms' baselines, but on some datasets the relative AUPRC gain is modest (e.g., Avito 5.59 vs. DRL 5.25; Stack 7.66 vs. MCMTAD 6.83). Reporting statistical significance tests, e.g., paired tests across seeds, would be more informative than mean±std alone.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular derivation; reconstruction losses and anomaly-score fusion are self-contained, and the experimental caveats are validity risks rather than definitional circularity.

full rationale

The claimed derivation chain is the reconstruction objective Eq. (17) and the anomaly score Eq. (18). Eq. (17) is L = L_rec + λ_s L_sparse + Σ L^r_q, where L_rec (Eq. 9) is mean squared reconstruction error of attribute blocks, L_sparse (Eq. 7) is a gate sparsity penalty, and L^r_q (Eq. 15) is a softplus edge-reconstruction loss over observed and sampled negative edges. Eq. (18) is a convex combination of normalized block-level and relation-level reconstruction residuals. No quantity in these equations is defined in terms of the final anomaly score, and no fitted parameter is a renamed label or target: the gates (Eqs. 3–5), encoders, decoders, and edge scorers are trained without anomaly labels, and labels enter only at evaluation. The authors' self-citations are contextual (related work, benchmark protocols) and are not load-bearing for the method's design; no uniqueness theorem or prior ansatz is invoked to force RelAD's choices. The benchmark injections in Appendix D do encode localized attribute deviations and relation redirections that match RelAD's inductive biases, and Appendix E's random hyperparameter search does not explicitly state a validation split; these are external-validity and possible label-leakage concerns, not circularity in the derivation. The paper's own stated limitation (no temporal modeling, Sec. H) is also not a circular step. Thus the framework is self-contained against its equations.

Assumptions & free parameters 5 free parameters · 6 assumptions · 0 invented entities

The central claim rests primarily on domain assumptions about how anomalies appear in relational data and on per-dataset benchmark construction, rather than on a formal derivation. No new physical or conceptual entities are postulated. The key uncontrolled choice is the injection design, which aligns with the model's two reconstruction modules.

free parameters (5)
  • α (attribute vs relational fusion weight) = tuned per dataset; values not reported
    Eq. (18); chosen by random search per dataset in Appendix E; Fig. 3 and Fig. 5 show strong sensitivity to α.
  • β (self vs child relational view weight) = tuned per dataset; values not reported
    Eq. (18); balances the two relational branches; tuned by random search, no fixed default.
  • λ_s (sparsity regularization coefficient) = tuned; not reported
    Eq. (17); controls how aggressively masks are sparsified; listed as a tuned hyperparameter in Appendix E.
  • Top-K (block and relation selection count) = not reported
    Eqs. (11) and (16); controls how many local signals survive the top-K aggregation; value is not given in the paper.
  • Anomaly injection parameters = 5% target; per-dataset values in Appendix D
    Injection rate, replacement counts, beacon-paper counts, cartel sizes, and redirect percentages define the benchmark labels. The central empirical claim depends on these choices.
assumptions (6)
  • domain assumption Reconstruction error is a valid anomaly signal: normal entities dominate and are reconstructible, anomalies deviate.
    Stated in Sec. 3 opening; this is the core inductive bias of reconstruction-based AD, not independently verified on real relational anomalies.
  • domain assumption Anomalies manifest as sparse local deviations concentrated in a few attribute blocks and/or relation-specific edges.
    Motivates top-K block selection (Eq. 11) and top-K relation selection (Eq. 16); the injection rules in Appendix D are designed to satisfy this shape.
  • domain assumption RelBench-style row encoding and child-table statistics (mean/std/count) preserve anomaly-relevant information.
    Sec. 2 'Initial Features' and Appendix E preprocessing assume aggregation over child tables does not destroy the anomaly signals the model later reconstructs.
  • domain assumption The central table and its foreign-key relations are correctly identified and define all relevant anomaly surfaces.
    Sec. 2 RAD setting assumes anomalies are about target entities and their typed relations, not table-wide drift or temporal events.
  • ad hoc to paper Uniform negative sampling for edge reconstruction produces useful contrast for relation-specific likelihoods.
    Eq. (15) and Algorithm 1 line 15 sample negative neighbors uniformly; no theoretical guarantee that this contrast separates anomalous edges.
  • standard math Standard neural-network training (MLPs, Adam, ReLU, layer norm) behaves as expected.
    Appendix E; unproved background required by the implementation, not specific to the relational anomaly claim.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Towards Anomaly Detection on Relational Data." pith.science (2026). https://pith.science/paper/MHLZJ7PI

@misc{pith2026260618621,
  author       = {Pith},
  title        = {Pith review of: Towards Anomaly Detection on Relational Data},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/MHLZJ7PI}},
  note         = {Machine review of arXiv:2606.18621}
}
read the original abstract

Relational databases are widely used for managing structured data in real-world systems. Detecting anomalies from such relational data is crucial for identifying fraud, risks, and abnormal behaviors, yet remains under-explored. The key challenges lie in the intrinsic complexity of relational data: multi-table attributes are high-dimensional and heterogeneous, making sparse abnormal clues easy to overwhelm by normal or irrelevant information. Moreover, anomalies may further manifest as abnormal connection patterns across different foreign-key relations, which existing tabular and graph anomaly detection methods are ill-suited to capture. To address them, we propose RelAD, a reconstruction-based framework that captures anomalies from both attribute and relational edge reconstruction. RelAD contains two core modules: conditional sparse-gated attribute reconstruction, which suppresses redundant multi-table attributes and emphasizes abnormal semantic blocks, and dual-view multi-relational edge reconstruction, which detects relation-specific abnormal connections from both intrinsic and behavioral instance profiles. The resulting attribute and relational signals are integrated through a lightweight fusion module to produce the final anomaly score. We further construct 6 benchmark datasets with systematic anomalies, on which extensive experiments show that RelAD consistently outperforms baselines while achieving competitive efficiency..The source code is available at https://github.com/Shiy-Li/RelAD.

Figures

Figures reproduced from arXiv: 2606.18621 by the authors.

Figure 1
Figure 1. Sketch maps of (a) Anomalous entity in relational database; and the limitations of adapting [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. The overall pipeline of RelAD for relational data anomaly detection. [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Sensitivity w.r.t. α and β. For β, most datasets prefer moderate values, confirm￾ing that both relational views provide complementary ev￾idence. This disparity suggests that anomalies in different relational databases may manifest primarily in attributes or in relational connections, and RelAD can flexibly adapt the fusion weights to the underlying anomaly source. Efficiency Analysis. To assess the runtime efficienc… view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Runtime comparison on the HM dataset. In this section, we briefly review three lines of related studies. A detailed literature review is in Appendix G. Relational Deep Learning (RDL) aims to learn from multi-table relational databases by model￾ing entities and their pr…
Figure 5
Figure 5. Figure 5: Sensitivity analysis of RelAD with respect to [PITH_FULL_IMAGE:figures/full_fig_p021_5.png]

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. CORE: In-Context Reconstruction for Unified Tabular Anomaly Detection

    cs.AI 2026-07 conditional novelty 6.0 of 10

    CORE detects anomalies in new tabular datasets by reconstructing each test sample from the nearest normal context samples in a learned, feature-aligned space; it is proposed as the first reconstruction-based unified t...

Reference graph

Works this paper leans on

60 extracted references · 3 linked inside Pith · cited by 1 Pith paper

  1. [1]

    Relbench: A benchmark for deep learning on relational databases.Advances in Neural Information Processing Systems, 37:21330–21341, 2024

    Joshua Robinson, Rishabh Ranjan, Weihua Hu, Kexin Huang, Jiaqi Han, Alejandro Dobles, Matthias Fey, Jan E Lenssen, Yiwen Yuan, Zecheng Zhang, et al. Relbench: A benchmark for deep learning on relational databases.Advances in Neural Information Processing Systems, 37:21330–21341, 2024

    2024

  2. [2]

    Relational deep learning: Challenges, foundations and next-generation architectures

    Vijay Prakash Dwivedi, Charilaos Kanatsoulis, Shenyang Huang, and Jure Leskovec. Relational deep learning: Challenges, foundations and next-generation architectures. InProceedings of the 31st ACM SIGKDD Conference on Knowledge Discovery and Data Mining V . 2, pages 5999–6009, 2025

    2025

  3. [3]

    Relational graph transformer

    Vijay Prakash Dwivedi, Sri Jaladi, Yangyi Shen, Federico Lopez, Charilaos I Kanatsoulis, Rishi Puri, Matthias Fey, and Jure Leskovec. Relational graph transformer. InTemporal Graph Learning Workshop@ KDD 2025

    2025

  4. [4]

    Relgnn: Composite message passing for relational deep learning

    Tianlang Chen, Charilaos Kanatsoulis, and Jure Leskovec. Relgnn: Composite message passing for relational deep learning. InInternational Conference on Machine Learning, pages 8296–8312. PMLR, 2025

    2025

  5. [5]

    Relational transformer: Toward zero-shot foundation models for relational data

    Rishabh Ranjan, Valter Hudovernik, Mark Znidar, Charilaos I Kanatsoulis, Roshan Reddy Upendra, Mahmoud Mohammadi, Joe Meyer, Tom Palczewski, Carlos Guestrin, and Jure Leskovec. Relational transformer: Toward zero-shot foundation models for relational data. InEurIPS 2025 Workshop: AI for Tabular Data

    2025

  6. [6]

    Griffin: Towards a graph-centric relational database foundation model

    Yanbo Wang, Xiyuan Wang, Quan Gan, Minjie Wang, Qibin Yang, David Wipf, and Muhan Zhang. Griffin: Towards a graph-centric relational database foundation model. InInternational Conference on Machine Learning, pages 64604–64627. PMLR, 2025

    2025

  7. [7]

    Beyond individual input for deep anomaly detection on tabular data

    Hugo Thimonier, Fabrice Popineau, Arpad Rimmel, and Bich-Liên Doan. Beyond individual input for deep anomaly detection on tabular data. InInternational Conference on Machine Learning, pages 48097–48123. PMLR, 2024

    2024

  8. [8]

    Drl: Decomposed representation learning for tabular anomaly detection

    Hangting Ye, He Zhao, Wei Fan, Mingyuan Zhou, Dan dan Guo, and Yi Chang. Drl: Decomposed representation learning for tabular anomaly detection. InThe Thirteenth International Conference on Learning Representations, 2025

    2025

Show all 60 references
  1. [9]

    Correcting false alarms from unseen: Adapting graph anomaly detectors at test time

    Junjun Pan, Yixin Liu, Chuan Zhou, Fei Xiong, Alan Wee-Chung Liew, and Shirui Pan. Correcting false alarms from unseen: Adapting graph anomaly detectors at test time. InProceedings of the AAAI Conference on Artificial Intelligence, 2026

    2026

  2. [10]

    A survey of generalization of graph anomaly detection: From transfer learning to foundation models

    Junjun Pan, Yu Zheng, Yue Tan, and Yixin Liu. A survey of generalization of graph anomaly detection: From transfer learning to foundation models. InThe 16th IEEE International Conference on Knowledge Graphs, 2025. 10

    2025

  3. [11]

    Deep anomaly detection on attributed networks

    Kaize Ding, Jundong Li, Rohit Bhanushali, and Huan Liu. Deep anomaly detection on attributed networks. InProceedings of the 2019 SIAM international conference on data mining, pages 594–602. SIAM, 2019

    2019

  4. [12]

    Anomaly detection on attributed networks via contrastive self-supervised learning.IEEE transactions on neural networks and learning systems, 33(6):2378–2392, 2021

    Yixin Liu, Zhao Li, Shirui Pan, Chen Gong, Chuan Zhou, and George Karypis. Anomaly detection on attributed networks via contrastive self-supervised learning.IEEE transactions on neural networks and learning systems, 33(6):2378–2392, 2021

    2021

  5. [13]

    Boosting graph anomaly detection with adaptive message passing

    Jingyan Chen, Guanghui Zhu, Chunfeng Yuan, and Yihua Huang. Boosting graph anomaly detection with adaptive message passing. InThe Twelfth International Conference on Learning Representations, 2024

    2024

  6. [14]

    Camera: Adapting to semantic camouflage in unsupervised text-attributed graph fraud detection

    Junjun Pan, Yixin Liu, Yu Zheng, Lianhua Chi, Alan Wee-Chung Liew, and Shirui Pan. Camera: Adapting to semantic camouflage in unsupervised text-attributed graph fraud detection. InInternational Joint Conference on Artificial Intelligence, 2026

    2026

  7. [15]

    Lof: identifying density-based local outliers

    Markus M Breunig, Hans-Peter Kriegel, Raymond T Ng, and Jörg Sander. Lof: identifying density-based local outliers. InProceedings of the 2000 ACM SIGMOD international conference on Management of data, pages 93–104, 2000

    2000

  8. [16]

    Lunar: Unifying local outlier detection methods via graph neural networks

    Adam Goodge, Bryan Hooi, See-Kiong Ng, and Wee Siong Ng. Lunar: Unifying local outlier detection methods via graph neural networks. InProceedings of the AAAI conference on artificial intelligence, volume 36, pages 6737–6745, 2022

    2022

  9. [17]

    Mcm: Masked cell modeling for anomaly detection in tabular data

    Jiaxin Yin, Yuanyuan Qiao, Zitang Zhou, Xiangchao Wang, and Jie Yang. Mcm: Masked cell modeling for anomaly detection in tabular data. InThe Twelfth International Conference on Learning Representations, 2024

    2024

  10. [18]

    Uncertainty- aware graph neural networks: A multihop evidence fusion approach.IEEE Transactions on Neural Networks and Learning Systems, 2025

    Qingfeng Chen, Shiyuan Li, Yixin Liu, Shirui Pan, Geoffrey I Webb, and Shichao Zhang. Uncertainty- aware graph neural networks: A multihop evidence fusion approach.IEEE Transactions on Neural Networks and Learning Systems, 2025

    2025

  11. [19]

    Influence-oriented personalized federated learning.arXiv preprint arXiv:2410.03315, 2024

    Yue Tan, Guodong Long, Jing Jiang, and Chengqi Zhang. Influence-oriented personalized federated learning.arXiv preprint arXiv:2410.03315, 2024

    2024 arXiv

  12. [20]

    Arc: A generalist graph anomaly detector with in-context learning.Advances in Neural Information Processing Systems, 37:50772–50804, 2024

    Yixin Liu, Shiyuan Li, Yu Zheng, Qingfeng Chen, Chengqi Zhang, and Shirui Pan. Arc: A generalist graph anomaly detector with in-context learning.Advances in Neural Information Processing Systems, 37:50772–50804, 2024

    2024

  13. [21]

    Freegad: A training- free yet effective approach for graph anomaly detection

    Yunfeng Zhao, Yixin Liu, Shiyuan Li, Qingfeng Chen, Yu Zheng, and Shirui Pan. Freegad: A training- free yet effective approach for graph anomaly detection. InProceedings of the 34th ACM International Conference on Information and Knowledge Management, pages 4379–4389, 2025

    2025

  14. [22]

    Relbench v2: A large-scale benchmark and repository for relational data.arXiv preprint arXiv:2602.12606, 2026

    Justin Gu, Rishabh Ranjan, Charilaos Kanatsoulis, Haiming Tang, Martin Jurkovic, Valter Hudovernik, Mark Znidar, Pranshu Chaturvedi, Parth Shroff, Fengyu Li, et al. Relbench v2: A large-scale benchmark and repository for relational data.arXiv preprint arXiv:2602.12606, 2026

    2026 arXiv

  15. [23]

    Salt: Sales autocompletion linked business tables dataset

    Tassilo Klein, Clemens Biehl, Margarida Costa, Andre Sres, Jonas Kolk, and Johannes Hoffart. Salt: Sales autocompletion linked business tables dataset. InNeurIPS 2024 Third Table Representation Learning Workshop

    2024

  16. [24]

    Prem: A simple yet effective approach for node- level graph anomaly detection

    Junjun Pan, Yixin Liu, Yizhen Zheng, and Shirui Pan. Prem: A simple yet effective approach for node- level graph anomaly detection. In2023 IEEE International Conference on Data Mining (ICDM), pages 1253–1258. IEEE, 2023

    2023

  17. [25]

    Efficient algorithms for mining outliers from large data sets

    Sridhar Ramaswamy, Rajeev Rastogi, and Kyuseok Shim. Efficient algorithms for mining outliers from large data sets. InProceedings of the 2000 ACM SIGMOD international conference on Management of data, pages 427–438, 2000

    2000

  18. [26]

    Generative adversarial active learning for unsupervised outlier detection.IEEE Transactions on Knowledge and Data Engineering, 32(8):1517–1528, 2019

    Yezheng Liu, Zhe Li, Chong Zhou, Yuanchun Jiang, Jianshan Sun, Meng Wang, and Xiangnan He. Generative adversarial active learning for unsupervised outlier detection.IEEE Transactions on Knowledge and Data Engineering, 32(8):1517–1528, 2019

    2019

  19. [27]

    Relational deep learning: Graph representation learning on relational databases

    Joshua Robinson, Rishabh Ranjan, Weihua Hu, Kexin Huang, Jiaqi Han, Alejandro Dobles, Matthias Fey, Jan Eric Lenssen, Yiwen Yuan, Zecheng Zhang, et al. Relational deep learning: Graph representation learning on relational databases. InNeurIPS 2024 third table representation le...

    2024

  20. [28]

    Large language models are good relational learners

    Fang Wu, Vijay Prakash Dwivedi, and Jure Leskovec. Large language models are good relational learners. InProceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 7835–7854, 2025. 11

    2025

  21. [29]

    Plurel: Synthetic data unlocks scaling laws for relational foundation models

    Vignesh Kothapalli, Rishabh Ranjan, Valter Hudovernik, Vijay Prakash Dwivedi, Johannes Hoffart, Carlos Guestrin, and Jure Leskovec. Plurel: Synthetic data unlocks scaling laws for relational foundation models. arXiv preprint arXiv:2602.04029, 2026

    2026 arXiv

  22. [30]

    Deep learning for anomaly detection: A review.ACM computing surveys (CSUR), 54(2):1–38, 2021

    Guansong Pang, Chunhua Shen, Longbing Cao, and Anton Van Den Hengel. Deep learning for anomaly detection: A review.ACM computing surveys (CSUR), 54(2):1–38, 2021

    2021

  23. [31]

    Deep neural networks and tabular data: A survey.IEEE transactions on neural networks and learning systems, 35(6):7499–7519, 2022

    Vadim Borisov, Tobias Leemann, Kathrin Seßler, Johannes Haug, Martin Pawelczyk, and Gjergji Kasneci. Deep neural networks and tabular data: A survey.IEEE transactions on neural networks and learning systems, 35(6):7499–7519, 2022

    2022

  24. [32]

    Anomaly detection for tabular data with internal contrastive learning

    Tom Shenkar and Lior Wolf. Anomaly detection for tabular data with internal contrastive learning. In International conference on learning representations, 2022

    2022

  25. [33]

    A comprehensive survey on graph anomaly detection with deep learning.IEEE transactions on knowledge and data engineering, 35(12):12012–12038, 2021

    Xiaoxiao Ma, Jia Wu, Shan Xue, Jian Yang, Chuan Zhou, Quan Z Sheng, Hui Xiong, and Leman Akoglu. A comprehensive survey on graph anomaly detection with deep learning.IEEE transactions on knowledge and data engineering, 35(12):12012–12038, 2021

    2021

  26. [34]

    Deep graph anomaly detection: A survey and new perspectives.IEEE Transactions on Knowledge and Data Engineer- ing, 2025

    Hezhe Qiao, Hanghang Tong, Bo An, Irwin King, Charu Aggarwal, and Guansong Pang. Deep graph anomaly detection: A survey and new perspectives.IEEE Transactions on Knowledge and Data Engineer- ing, 2025

    2025

  27. [35]

    From few-shot to zero-shot: Towards generalist graph anomaly detection.IEEE Transactions on Knowledge and Data Engineering, 2026

    Yixin Liu, Shiyuan Li, Yu Zheng, Qingfeng Chen, Chengqi Zhang, Philip S Yu, and Shirui Pan. From few-shot to zero-shot: Towards generalist graph anomaly detection.IEEE Transactions on Knowledge and Data Engineering, 2026

    2026

  28. [36]

    Raising the bar in graph ood generalization: Invariant learning beyond explicit environment modeling.IEEE Transactions on Pattern Analysis and Machine Intelligence, 2026

    Xu Shen, Yixin Liu, Yili Wang, Rui Miao, Yiwei Dai, Shirui Pan, Yi Chang, and Xin Wang. Raising the bar in graph ood generalization: Invariant learning beyond explicit environment modeling.IEEE Transactions on Pattern Analysis and Machine Intelligence, 2026

    2026

  29. [37]

    Fedcigar: A personalized reconstruction approach for federated graph-level anomaly detection

    Yunfeng Zhao, Yixin Liu, Qingfeng Chen, Shiyuan Li, Yue Tan, and Shirui Pan. Fedcigar: A personalized reconstruction approach for federated graph-level anomaly detection. InInternational Joint Conference on Artificial Intelligence, 2026

    2026

  30. [38]

    Rethinking feature alignment in generalist graph anomaly detection: A relational fingerprint-based approach

    Yujing Liu, Yixin Liu, Yu Zheng, Alan Wee-Chung Liew, Xiaofeng Cao, and Shirui Pan. Rethinking feature alignment in generalist graph anomaly detection: A relational fingerprint-based approach. In International Conference on Machine Learning, 2026

    2026

  31. [39]

    Anomalous: A joint modeling approach for anomaly detection on attributed networks

    Zhen Peng, Minnan Luo, Jundong Li, Huan Liu, Qinghua Zheng, et al. Anomalous: A joint modeling approach for anomaly detection on attributed networks. InIjcai, volume 18, pages 3513–3519, 2018

    2018

  32. [40]

    Radar: residual analysis for anomaly detection in attributed networks

    Jundong Li, Harsh Dani, Xia Hu, and Huan Liu. Radar: residual analysis for anomaly detection in attributed networks. InProceedings of the 26th International Joint Conference on Artificial Intelligence, pages 2152–2158, 2017

    2017

  33. [41]

    Truncated affinity maximization: One-class homophily modeling for graph anomaly detection.Advances in Neural Information Processing Systems, 36:49490–49512, 2023

    Hezhe Qiao and Guansong Pang. Truncated affinity maximization: One-class homophily modeling for graph anomaly detection.Advances in Neural Information Processing Systems, 36:49490–49512, 2023

    2023

  34. [42]

    Modeling relational data with graph convolutional networks

    Michael Schlichtkrull, Thomas N Kipf, Peter Bloem, Rianne Van Den Berg, Ivan Titov, and Max Welling. Modeling relational data with graph convolutional networks. InEuropean semantic web conference, pages 593–607. Springer, 2018

    2018

  35. [43]

    Blindguard: Safeguarding llm-based multi-agent systems under unknown attacks

    Rui Miao, Yixin Liu, Yili Wang, Xu Shen, Yue Tan, Yiwei Dai, Shirui Pan, and Xin Wang. Blindguard: Safeguarding llm-based multi-agent systems under unknown attacks. InProceedings of the 64th Annual Meeting of the Association for Computational Linguistics, 2026

    2026

  36. [44]

    Dynhd: Hallucination detection for diffusion large language models via denoising dynamics deviation learning.arXiv preprint arXiv:2603.16459, 2026

    Yanyu Qian, Yue Tan, Yixin Liu, Wang Yu, and Shirui Pan. Dynhd: Hallucination detection for diffusion large language models via denoising dynamics deviation learning.arXiv preprint arXiv:2603.16459, 2026

    2026

  37. [45]

    Assemble your crew: Automatic multi-agent communication topology design via autoregressive graph generation

    Shiyuan Li, Yixin Liu, Qingsong Wen, Chengqi Zhang, and Shirui Pan. Assemble your crew: Automatic multi-agent communication topology design via autoregressive graph generation. InProceedings of the AAAI Conference on Artificial Intelligence, volume 40, pages 23142–23150, 2026

    2026

  38. [46]

    Ofa-mas: One-for-all multi-agent system topology design based on mixture-of-experts graph generative models

    Shiyuan Li, Yixin Liu, Yu Zheng, Mei Li, Quoc Viet Hung Nguyen, and Shirui Pan. Ofa-mas: One-for-all multi-agent system topology design based on mixture-of-experts graph generative models. InProceedings of the ACM Web Conference 2026, pages 1333–1344, 2026

    2026

  39. [47]

    Rethinking unsupervised time series anomaly detection: Dynamic attention based on route inverse-masking.Applied Soft Computing, page 113971, 2025

    Enguang Zuo, Jie Zhong, Chen Chen, Cheng Chen, Kurban Ubul, and Xiaoyi Lv. Rethinking unsupervised time series anomaly detection: Dynamic attention based on route inverse-masking.Applied Soft Computing, page 113971, 2025. 12

    2025

  40. [48]

    Enguang Zuo, Junyi Yan, Alimjan Aysa, Chen Chen, Cheng Chen, Hongbing Ma, Xiaoyi Lv, and Kurban Ubul. Sucola: Self-adaptive structure refinement unsupervised contrastive learning framework for food safety risk early warning.Engineering Applications of Artificial Intelligence, ...

    2023

  41. [49]

    Retrieval augmented deep anomaly detection for tabular data

    Hugo Thimonier, Fabrice Popineau, Arpad Rimmel, and Bich-Liên Doan. Retrieval augmented deep anomaly detection for tabular data. InProceedings of the 33rd ACM international conference on information and knowledge management, pages 2250–2259, 2024

    2024

  42. [50]

    Anollm: Large language models for tabular anomaly detection

    Che-Ping Tsai, Ganyu Teng, Phillip Wallis, and Wei Ding. Anollm: Large language models for tabular anomaly detection. InThe Thirteenth International Conference on Learning Representations, 2025

    2025

  43. [51]

    Disentangling tabular data towards better one-class anomaly detection

    Jianan Ye, Zhaorui Tan, Yijie Hu, Xi Yang, Guangliang Cheng, and Kaizhu Huang. Disentangling tabular data towards better one-class anomaly detection. InProceedings of the AAAI Conference on Artificial Intelligence, volume 39, pages 13061–13068, 2025

    2025

  44. [52]

    Data-efficient and interpretable tabular anomaly detection

    Chun-Hao Chang, Jinsung Yoon, Sercan Ö Arik, Madeleine Udell, and Tomas Pfister. Data-efficient and interpretable tabular anomaly detection. InProceedings of the 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, pages 190–201, 2023

    2023

  45. [53]

    Towards one-for-all anomaly detection for tabular data

    Shiyuan Li, Yixin Liu, Yu Zheng, Xiaofeng Cao, Shirui Pan, and Heng Tao Shen. Towards one-for-all anomaly detection for tabular data. InForty-third international conference on machine learning, 2026

    2026

  46. [54]

    Graph anomaly detection in time series: A survey.IEEE Transactions on Pattern Analysis and Machine Intelligence, 2025

    Thi Kieu Khanh Ho, Ali Karami, and Narges Armanfard. Graph anomaly detection in time series: A survey.IEEE Transactions on Pattern Analysis and Machine Intelligence, 2025

    2025

  47. [55]

    Address anomalies at critical crossroads for graph anomaly detection.IEEE Transactions on Knowledge and Data Engineering, 2025

    Junyi Yan, Enguang Zuo, Ke Liang, Meng Liu, Miaomiao Li, Xinwang Liu, Xiaoyi Lv, and Kai Lu. Address anomalies at critical crossroads for graph anomaly detection.IEEE Transactions on Knowledge and Data Engineering, 2025

    2025

  48. [56]

    Smoothgnn: Smoothing-aware gnn for unsupervised node anomaly detection

    Xiangyu Dong, Xingyi Zhang, Yanni Sun, Lei Chen, Mingxuan Yuan, and Sibo Wang. Smoothgnn: Smoothing-aware gnn for unsupervised node anomaly detection. InProceedings of the ACM on Web Conference 2025, pages 1225–1236, 2025

    2025

  49. [57]

    verified purchase

    Rui Bing, Guan Yuan, Mu Zhu, Fanrong Meng, Huifang Ma, and Shaojie Qiao. Heterogeneous graph neural networks analysis: a survey of techniques, evaluations and applications.Artificial Intelligence Review, 56(8), 2023. 13 A Algorithm Description The training and testing algorith...

    2023

  50. [58]

    Importantly, anomaly labels are assigned only after verifying that the sampled entities satisfy the dataset-specific injection constraints

    Stratified Injection Rate:We initially sample 5% of the total population across all datasets using a stratified sampling strategy (e.g., binning by node activity or degree) to eliminate selection bias, ensuring that the statistical distribution of the anomalous group rigorousl...

  51. [59]

    Replace-Only

    Strict “Replace-Only” Strategy:We strictly enforce a “replace-only” strategy during the injection phase, firmly prohibiting the addition or deletion of any data rows. This guarantees that the foundational statistical features of the nodes (e.g., interaction frequency, total de...

  52. [60]

    Guided by these principles, we design a separate injection rule for each dataset according to its specific schema and business context

    Real-World Scenario Reconstruction:Each injection strategy is explicitly designed to reconstruct real-world business fraud scenarios, discarding mere random noise. Guided by these principles, we design a separate injection rule for each dataset according to its specific schema...

Pith tools

Reviewed August 4, 2026 · model on record in the stance chip above.