Page 131 - 《软件学报》2026年第4期

1572 软件学报 2026 年第 37 卷第 4 期

线性 Transformer 高效地建模全局节点表征, 并在全局和局部视角下将节点表示转换为自身及邻居之间的残差形
式以生成数据集无关的通用特征, 最后结合少样本重建学习实现无需目标数据微调的高效检测. 实验表明, GRAD
在跨领域基准数据集上优于现有方法. 在本文, 可以发现, 所计算出的残差特征可视为对一阶信息的利用, 所以对
于多阶残差信息的有效融合是未来值得探索的内容. 同时, 对于线性 Transformer, 它仍是基于多层神经网络进行
建模, 这扩大了模型规模, 在异常检测这种数据量少的场景下, 增加了模型过拟合的风险. 因此, 利用数据增广技术
来扩充训练数据也是未来的研究方向之一.

References
[1] Wu ZH, Pan SR, Chen FW, Long GD, Zhang CQ, Yu PS. A comprehensive survey on graph neural networks. IEEE Trans. on Neural
Networks and Learning Systems, 2021, 32(1): 4–24. [doi: 10.1109/TNNLS.2020.2978386]
[2] Qiao HZ, Tong HH, An B, King I, Aggarwal C, Pang GS. Deep graph anomaly detection: A survey and new perspectives. IEEE Trans. on
Knowledge and Data Engineering, 2025, 37(9): 5106–5126. [doi: 10.1109/TKDE.2025.3581578]
[3] Pan JJ, Liu YX, Zheng X, Zheng YZ, Liew AWC, Li FY, Pan SR. A label-free heterophily-guided approach for unsupervised graph fraud
detection. In: Proc. of the 39th AAAI Conf. on Artificial Intelligence. Philadelphia: AAAI Press, 2025. 12443–12451. [doi: 10.1609/aaai.
v39i12.33356]
[4] Li JD, Dani H, Hu X, Liu H. Radar: Residual analysis for anomaly detection in attributed networks. In: Proc. of the 26th Int’l Joint Conf.
on Artificial Intelligence. Melbourne: AAAI Press, 2017. 2152–2158.
[5] Erfani SM, Rajasegarar S, Karunasekera S, Leckie C. High-dimensional and large-scale anomaly detection using a linear one-class SVM
with deep learning. Pattern Recognition, 2016, 58: 121–134. [doi: 10.1016/j.patcog.2016.03.028]
[6] Kipf TN, Welling M. Semi-supervised classification with graph convolutional networks. arXiv:1609.02907, 2017.
[7] Xiao SX, Wang SP, Dai YF, Guo WZ. Graph neural networks in node classification: Survey and evaluation. Machine Vision and
Applications, 2022, 33(1): 4. [doi: 10.1007/s00138-021-01251-0]
[8] Zhang MH, Chen YX. Link prediction based on graph neural networks. In: Proc. of the 32nd Int’l Conf. on Neural Information
Processing Systems. Montréal: Curran Associates Inc., 2018. 5171–5181.
[9] Ma XX, Wu J, Xue S, Yang J, Zhou C, Sheng QZ, Xiong H, Akoglu L. A comprehensive survey on graph anomaly detection with deep
learning. IEEE Trans. on Knowledge and Data Engineering, 2023, 35(12): 12012–12038. [doi: 10.1109/TKDE.2021.3118815]
[10] Liu YX, Li SY, Zheng Y, Chen QF, Zhang CQ, Pan SR. ARC: A generalist graph anomaly detector with in-context learning. In: Proc. of
the 38th Int’l Conf. on Neural Information Processing Systems. Vancouver: Curran Associates Inc., 2025. 50772–50804.
[11] Qiao HZ, Niu CX, Chen L, Pang GS. AnomalyGFM: Graph foundation model for zero/few-shot anomaly detection. In: Proc. of the 31st
ACM SIGKDD Conf. on Knowledge Discovery and Data Mining V.2. Toronto: ACM, 2025. 2326–2337. [doi: 10.1145/3711896.
3736843]
[12] Zhu JW, Pang GS. Toward generalist anomaly detection via in-context residual learning with few-shot sample prompts. In: Proc. of the
2024 IEEE/CVF Conf. on Computer Vision and Pattern Recognition. Seattle: IEEE, 2024. 17826–17836. [doi: 10.1109/CVPR52733.2024.
01688]
[13] Zheng WQ, Huang EW, Rao N, Katariya S, Wang ZY, Subbian K. Cold brew: Distilling graph node representations with incomplete or
missing neighborhoods. arXiv:2111.04840, 2022.
[14] Zhang JQ, Wang SZ, Chen SC. Reconstruction enhanced multi-view contrastive learning for anomaly detection on attributed networks.
arXiv:2205.04816, 2022.
[15] Jin M, Liu YX, Zheng Y, Chi LH, Li YF, Pan SR. ANEMONE: Graph anomaly detection with multi-scale contrastive learning. In: Proc.
of the 30th Int’l Conf. on Information & Knowledge Management. Queensland: ACM, 2021. 3122–3126. [doi: 10.1145/3459637.
3482057]
[16] Tang JH, Li JJ, Gao ZQ, Li J. Rethinking graph neural networks for anomaly detection. In: Proc. of the 39th Int’l Conf. on Machine
Learning. Maryland: PMLR, 2022. 21076–21089.
[17] Wu QT, Zhao WT, Yang CX, Zhang HR, Nie F, Jiang HT, Bian YT, Yan JC. SGFormer: Simplifying and empowering Transformers for
large-graph representations. In: Proc. of the 37th Int’l Conf. on Neural Information Processing Systems. New Orleans: Curran Associates
Inc., 2023. 64753–64773.
[18] Wu QT, Zhao WT, Li Z, Wipf DP, Yan JC. Nodeformer: A scalable graph structure learning Transformer for node classification. In: Proc.
of the 36th Int’l Conf. on Neural Information Processing Systems. New Orleans: Curran Associates Inc., 2022. 27387–27401.
[19] Geisler S, Kosmala A, Herbst D, Günnemann S. Spatio-spectral graph neural networks. In: Proc. of the 38th Int’l Conf. on Neural