CoChat
Sign Up
Cited by 169Advanced Graph Neural NetworksAnomaly Detection Techniques and ApplicationsComplex Network Analysis Techniques

Anomaly Detection in Dynamic Graphs via Transformer

Open access
Liu, Yixin, Pan, Shirui, Wang, Yu Guang, Xiong, Fei, Wang, Liang, Chen, Qingfeng, Lee, Vincent CS·2023·IEEE Transactions on Knowledge and Data Engineering
DOI 10.1109/tkde.2021.3124061
Open this paper in CoChat — free accountOpen PDF

Abstract

Detecting anomalies for dynamic graphs has drawn increasing attention due to their wide applications in social networks, e-commerce, and cybersecurity. Recent deep learning-based approaches have shown promising results over shallow methods. However, they fail to address two core challenges of anomaly detection in dynamic graphs: the lack of informative encoding for unattributed nodes and the difficulty of learning discriminate knowledge from coupled spatial-temporal dynamic graphs. To overcome these challenges, in this paper, we present a novelTransformer-basedAnomalyDetection framework forDYnamic graphs (TADDY). Our framework constructs a comprehensive node encoding strategy to better represent each node’s structural and temporal roles in an evolving graphs stream. Meanwhile, TADDY captures informative representation from dynamic graphs with coupled spatial-temporal patterns via a dynamic graph transformer model. The extensive experimental results demonstrate that our proposed TADDY framework outperforms the state-of-the-art methods by a large margin on six real-world datasets.

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1109/tkde.2021.3124061

Cites

5 of 5 references verified0 citation errors · 0 not checkable

Statistics

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • These challenges have led to the development of graph neural frameworks that incorporate heterogeneous modeling and dynamic reasoning.

    “These challenges have led to the development of graph neural frameworks that incorporate heterogeneous modeling and dynamic reasoning. [1], [25], [26], [34]”

    From A Survey of Heterogeneous Graph Neural Networks for Cybersecurity Anomaly Detection (2025) · cites this paper as [1]

    The Introduction (p. 2) describes TADDY as a framework that learns coupled spatial and temporal information from dynamic graphs, supporting the dynamic-reasoning part of the claim. Section 2.2 (p. 3) discusses heterogeneous-graph transformers as separate prior works; it does not say that TADDY incorporates heterogeneous modeling alongside dynamic reasoning. (LLM, conf=0.86)

    Experimental — accuracy not yet measured

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Cites

Flagged first, then by citations

Research works printed but not linked: 45

  1. [1][1] L. Wang, Z. Yu, F. Xiong, D. Yang, S. Pan, and Z. Yan, 'Influence spread in geo-social networks: a multiobjective optimization perspective,' IEEE TCYB , 2019. -
  2. [3][3] Y. Zheng, R. Hu, S.-f. Fung, C. Yu, G. Long, T. Guo, and S. Pan, 'Clustering social audiences in business information networks,' Pattern Recognition , vol. 100, p. 107126, 2020. -
  3. [4][4] Y. Gao, X. Li, H. Peng, B. Fang, and P. Yu, 'Hincti: A cyber threat intelligence modeling and identification system based on heterogeneous information network,' IEEE TKDE , 2020. -
  4. [5][5] D. Jin, Z. Yu, P. Jiao, S. Pan, D. He, J. Wu, P. S. Yu, and W. Zhang, 'A survey of community detection approaches: From statistical modeling to deep learning,' IEEE TKDE , 2021. -
  5. [6][6] F. Xia, K. Sun, S. Yu, A. Aziz, L. Wan, S. Pan, and H. Liu, 'Graph learning: A survey,' IEEE Transactions on Artificial Intelligence , 2021. -
  6. [7][7] H. Peng, R. Yang, Z. Wang, J. Li, L. He, P. Yu, A. Zomaya, and R. Ranjan, 'Lime: Low-cost incremental learning for dynamic heterogeneous information networks,' IEEE Transactions on Computers , pp. 1-1, 2021. -
  7. [8][8] P. Jiao, X. Guo, X. Jing, D. He, H. Wu, S. Pan, M. Gong, and W. Wang, 'Temporal network embedding for link prediction via vae joint attention mechanism,' IEEE TNNLS , 2021. -
  8. [9][9] W. Yu, W. Cheng, C. C. Aggarwal, K. Zhang, H. Chen, and W. Wang, 'NetWalk: A flexible deep embedding approach for anomaly detection in dynamic networks,' in SIGKDD , 2018, pp. 2672-2681. -
  9. [10][10] L. Zheng, Z. Li, J. Li, Z. Li, and J. Gao, 'AddGraph: Anomaly detection in dynamic graph using attention-based temporal gcn.' in IJCAI , 2019, pp. 4419-4425. -
  10. [11][11] C. C. Aggarwal, Y. Zhao, and S. Y. Philip, 'Outlier detection in graph streams,' in ICDE . IEEE, 2011, pp. 399-409. -
  11. [12][12] S. Ranshous, S. Harenberg, K. Sharma, and N. F. Samatova, 'A scalable approach for outlier detection in edge streams using sketch-based approximations,' in SDM . SIAM, 2016, pp. 189197. -
  12. [13][13] L. Cai, Z. Chen, C. Luo, J. Gui, J. Ni, D. Li, and H. Chen, 'Structural temporal graph neural networks for anomaly detection in dynamic graphs,' arXiv preprint arXiv:2005.07427 , 2020. - · arXiv
  13. [14][14] C. Yang, L. Zhou, H. Wen, Z. Zhou, and Y. Wu, 'H-VGRAE: A hierarchical stochastic spatial-temporal embedding method for robust anomaly detection in dynamic networks,' arXiv preprint arXiv:2007.06903 , 2020. - · arXiv
  14. [15][15] H. Peng, J. Li, Y. Song, R. Yang, R. Ranjan, P. Yu, and L. He, 'Streaming social event detection and evolution discovery in heterogeneous information networks,' ACM TKDD , 2021. -
  15. [16][16] K. Sricharan and K. Das, 'Localizing anomalous changes in timeevolving graphs,' in SIGMOD/PODS , 2014, pp. 1347-1358. -
  16. [17][17] E. Manzoor, S. M. Milajerdi, and L. Akoglu, 'Fast memoryefficient anomaly detection in streaming heterogeneous graphs,' in SIGKDD , 2016, pp. 1035-1044. -
  17. [18][18] D. Eswaran, C. Faloutsos, S. Guha, and N. Mishra, 'Spotlight: Detecting anomalies in streaming graphs,' in SIGKDD , 2018, pp. 1378-1386. -
  18. [19][19] T. N. Kipf and M. Welling, 'Semi-supervised classification with graph convolutional networks,' in ICLR , 2017. -
  19. [20][20] A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A. N. Gomez, Ł. Kaiser, and I. Polosukhin, 'Attention is all you need,' in NeurIPS , 2017, pp. 6000-6010. -
  20. [22][22] Z. Lan, M. Chen, S. Goodman, K. Gimpel, P. Sharma, and R. Soricut, 'ALBERT: A lite bert for self-supervised learning of language representations,' in ICLR , 2019. -
  21. [23][23] Y. Liu, M. Ott, N. Goyal, J. Du, M. Joshi, D. Chen, O. Levy, M. Lewis, L. Zettlemoyer, and V. Stoyanov, 'RoBERTa: A robustly optimized bert pretraining approach,' arXiv preprint arXiv:1907.11692 , 2019. - · arXiv
  22. [24][24] Z. Yang, Z. Dai, Y. Yang, J. Carbonell, R. R. Salakhutdinov, and Q. V. Le, 'XLNet: Generalized autoregressive pretraining for language understanding,' NeurIPS , vol. 32, pp. 5753-5763, 2019. -
  23. [25][25] L. Liu, W. L. Hamilton, G. Long, J. Jiang, and H. Larochelle, 'A universal representation transformer layer for few-shot image classification,' in ICLR , 2021. -
  24. [26][26] N. Carion, F. Massa, G. Synnaeve, N. Usunier, A. Kirillov, and S. Zagoruyko, 'End-to-end object detection with transformers,' in ECCV . Springer, 2020, pp. 213-229. -
  25. [27][27] A. Dosovitskiy, L. Beyer, A. Kolesnikov, D. Weissenborn, X. Zhai, T. Unterthiner, M. Dehghani, M. Minderer, G. Heigold, S. Gelly, J. Uszkoreit, and N. Houlsby, 'An image is worth 16x16 words: Transformers for image recognition at scale,' in ICLR , 2021. -
  26. [29][29] K. Han, Y. Wang, H. Chen, X. Chen, J. Guo, Z. Liu, Y. Tang, A. Xiao, C. Xu, Y. Xu et al. , 'A survey on visual transformer,' arXiv preprint arXiv:2012.12556 , 2020. - · arXiv
  27. [30][30] Y. Tay, M. Dehghani, D. Bahri, and D. Metzler, 'Efficient transformers: A survey,' arXiv preprint arXiv:2009.06732 , 2020. - · arXiv
  28. [31][31] S. Yun, M. Jeong, R. Kim, J. Kang, and H. J. Kim, 'Graph transformer networks,' NeurIPS , vol. 32, pp. 11 983-11 993, 2019. -
  29. [32][32] Z. Hu, Y. Dong, K. Wang, and Y. Sun, 'Heterogeneous graph transformer,' in WWW , 2020, pp. 2704-2710. -
  30. [33][33] Y. Rong, Y. Bian, T. Xu, W. Xie, Y. Wei, W. Huang, and J. Huang, 'Self-supervised graph transformer on large-scale molecular data,' NeurIPS , vol. 33, 2020. -
  31. [34][34] J. Zhang, H. Zhang, C. Xia, and L. Sun, 'Graph-bert: Only attention is needed for learning graph representations,' arXiv preprint arXiv:2001.05140 , 2020. - · arXiv
  32. [35][35] Y. Liu, S. Pan, M. Jin, C. Zhou, F. Xia, and P. S. Yu, 'Graph selfsupervised learning: A survey,' arXiv preprint arXiv:2103.00111 , 2021. - · arXiv
  33. [36][36] Y. Liu, Z. Li, S. Pan, C. Gong, C. Zhou, and G. Karypis, 'Anomaly detection on attributed networks via contrastive self-supervised learning,' IEEE TNNLS , 2021. -
  34. [37][37] J. Klicpera, S. Weiß enberger, and S. G¨ unnemann, 'Diffusion improves graph learning,' in NeurIPS , vol. 32, 2019, pp. 13 33313 345. -
  35. [38][38] K. Hassani and A. H. Khasahmadi, 'Contrastive multi-view representation learning on graphs,' in ICML , 2020, pp. 3451-3461. -
  36. [40][40] R. I. Kondor and J. Lafferty, 'Diffusion kernels on graphs and other discrete structures,' in ICML , vol. 2002, 2002, pp. 315-322. -
  37. [42][42] T. Opsahl and P. Panzarasa, 'Clustering in weighted networks,' Social networks , vol. 31, no. 2, pp. 155-163, 2009. -
  38. [43][43] M. De Choudhury, H. Sundaram, A. John, and D. D. Seligmann, 'Social synchrony: Predicting mimicry of user actions in online social media,' in International conference on computational science and engineering , vol. 4. IEEE, 2009, pp. 151-158. -
  39. [44][44] R. A. Rossi and N. K. Ahmed, 'The network data repository with interactive graph analytics and visualization,' in AAAI , 2015. [Online]. Available: http://networkrepository.com -
  40. [45][45] S. Kumar, F. Spezzano, V. Subrahmanian, and C. Faloutsos, 'Edge weight prediction in weighted signed networks,' in ICDM . IEEE, 2016, pp. 221-230. -
  41. [46][46] S. Kumar, B. Hooi, D. Makhija, M. Kumar, C. Faloutsos, and V. Subrahmanian, 'Rev2: Fraudulent user prediction in rating platforms,' in WSDM . ACM, 2018, pp. 333-341. -
  42. [47][47] B. Zhang, R. Liu, D. Massey, and L. Zhang, 'Collecting the internet as-level topology,' ACM SIGCOMM Computer Communication Review , vol. 35, no. 1, pp. 53-61, 2005. -
  43. [48][48] B. Perozzi, R. Al-Rfou, and S. Skiena, 'Deepwalk: Online learning of social representations,' in SIGKDD , 2014, pp. 701-710. -
  44. [50][50] U. Von Luxburg, 'A tutorial on spectral clustering,' Statistics and computing , vol. 17, no. 4, pp. 395-416, 2007. -
  45. [51][51] A. Paszke, S. Gross, F. Massa, A. Lerer, J. Bradbury, G. Chanan, T. Killeen, Z. Lin, N. Gimelshein, L. Antiga et al. , 'Pytorch: An imperative style, high-performance deep learning library,' NeurIPS , vol. 32, pp. 8026-8037, 2019. <!-- image --> <!-- image --> <!-- image --> <!-- image --> <!-- image --> <!-- image --> <!-- image --> Yixin Liu received the B.S. degree and M.S. degree from Beihang University, Beijing, China, in 2017 and 2020, respectively. He is currently pursuing his Ph.D. degree in computer science at Monash University, Melbourne, Australia. His research concentrates on data mining, machine learning, and deep learning on graphs. Shirui Pan received a Ph.D. in computer science from the University of Technology Sydney (UTS), Ultimo, NSW, Australia. He is currently a Senior Lecturer with the Faculty of Information Technology, Monash University, Australia. He is an ARC Future Fellow (awarded in 2021). His research interests include data mining and machine learning. To date, Dr Pan has published over 100 research papers in top-tier journals and conferences, including TPAMI, TNNLS, and TKDE. Yu Guang Wang received a Ph.D. in applied mathematics from University of New South Wales, Australia. He is an adjunct associate lecturer at UNSW Sydney. He is also a scientist at Max Planck Institute for Mathematics in Sciences, in Mathematics Machine Learning group. His research interests lie in computational mathematics, statistics, machine learning, and data science. Fei Xiong received the Ph.D. degree from Beijing Jiaotong University, Beijing, China, in 2013. He was a Visiting Scholar with Carnegie Mellon University, Pittsburgh, PA, USA, from 2011 to 2012. He is currently an Associate Professor with the School of Electronic and Information Engineering, Beijing Jiaotong University. His current research interests include Web mining, complex networks, and complex systems. Liang Wang the Ph.D. degree in computer science from the Shenyang Institute of Automation, Chinese Academy of Sciences, Shenyang, China, in 2014. He was a Post-Doctoral Researcher with Northwestern Polytechnical University, Xi'an, China, in 2017, where he is currently an Associate Professor. His current research interests include ubiquitous computing, mobile crowd sensing, and data mining. Qingfeng Chen received the BSc and MSc degrees in mathematics from Guangxi Normal University, China, in 1995 and 1998, respectively, and the PhD degree in computer science from the University of Technology Sydney, in September 2004. He is now a professor with Guangxi University, China, and the Hundred Talent Program of Guangxi. His research interests include bioinformatics, data mining, and artificial intelligence. Vincent CS Lee received the PhD degree in adaptive systems from The University of NewCastle, Australia, in 1992. He is currently an associate Professor with the Department of Data Science and Artificial Intelligence, Faculty of IT, Monash University, Australia. He is a multiinterdisciplinary researcher spanning adaptive signal processing and control system, computational intelligence, AI in economic and finance, deep machine learning and computer vision, digital health process mining, and information secu- rity and network cryptography disciplines. He is a senior member of the IEEE.

Cited by

Sorted by citations

Authors
Liu, Y., Pan, S., Wang, Y. G., Xiong, F., Wang, L., Chen, Q., & Lee, V. C.
Affiliations
  • Monash University (AU)
  • Shanghai Jiao Tong University (CN)
  • Beijing Jiaotong University (CN)
  • Northwestern Polytechnical University (CN)
  • Guangxi University (CN)
ORCID
Venue
IEEE Transactions on Knowledge and Data Engineering
Publisher
Institute of Electrical and Electronics Engineers (IEEE)
Published
2023-12-01
Volume
35
Issue
12
Pages
12081-12094
ISSN
  • 1041-4347 (ISSN-L)
  • 1558-2191
  • 2326-3865
DOI
10.1109/tkde.2021.3124061
OpenAlex
W3191802839
Cited by
169
References
5
Citations by year
2026: 52 · 2025: 57 · 2024: 29 · 2023: 22 · 2022: 8 · 2021: 1
FWCI
5.05
Type
journal article
Language
en
Open access
Licence
Locations
  • IEEE Transactions on Knowledge and Data Engineering (publishedVersion)
  • arXiv (Cornell University) (submittedVersion)
  • Griffith Research Online (submittedVersion)
  • arXiv (Cornell University) (cc-by)
Funders
  • Australian Research Council (FT210100097)
  • National Natural Science Foundation of China (61963004)
Topics
  • Physical Sciences › Computer Science › Artificial Intelligence › Advanced Graph Neural Networks
  • Physical Sciences › Computer Science › Artificial Intelligence › Anomaly Detection Techniques and Applications
  • Physical Sciences › Physics and Astronomy › Statistical and Nonlinear Physics › Complex Network Analysis Techniques
Keywords
anomaly detection, dynamic graphs, graph transformer, spatial-temporal patterns

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review