{"type":"ScholarlyArticle","title":"Unsupervised Anomaly Detection in Dynamic Graphs via Compatibility Modeling and Boundary Learning","authors":[{"name":"Jiachi Luo","slug":"jiachi-luo","given":"Jiachi","family":"Luo","affiliation":""},{"name":"Shameng Wen","slug":"shameng-wen","given":"Shameng","family":"Wen","affiliation":""},{"name":"Ziyang Qiu","slug":"ziyang-qiu","given":"Ziyang","family":"Qiu","affiliation":""},{"name":"Chen Jiang","slug":"chen-jiang","given":"Chen","family":"Jiang","affiliation":""},{"name":"Qi Huang","slug":"qi-huang","given":"Qi","family":"Huang","affiliation":""},{"name":"Yuxing Tian","slug":"yuxing-tian","given":"Yuxing","family":"Tian","affiliation":""},{"name":"Aiwen Jiang","slug":"aiwen-jiang","given":"Aiwen","family":"Jiang","affiliation":""}],"abstract":"Anomaly detection in dynamic graphs is essential for monitoring evolving systems such as transaction networks and online platforms. Yet existing methods remain limited in realistic edge-stream settings: snapshot-based approaches discretize continuous interactions and miss fine-grained temporal signals, while many continuous-time models rely on scarce node/edge attributes and often fail to explicitly assess whether a destination is compatible with a source’s recent context. Moreover, under extreme class imbalance and the lack of anomaly labels, unsupervised detectors frequently yield ill-defined normality criteria and ambiguous decision boundaries. We propose BAD, an unsupervised framework for anomaly detection in continuous-time dynamic graphs. BAD adopts a minimalistic design that represents nodes with learnable identity embeddings and performs pairwise compatibility modeling via cross-attention between each destination node and the source’s recent neighbors, enabling direct characterization of context-dependent deviations without requiring attributes. To obtain a principled separating boundary, BAD further integrates normalizing flows to model the distribution of normal interactions and derive likelihood-based anomaly scores. Extensive experiments on four real-world datasets demonstrate that BAD consistently performs well, highlighting its robustness under feature-scarce and label-scarce conditions.","doi":"10.24963/ijcai.2026/322","url":"https://ijcai.org/proceedings/2026/322","pdf":"https://ijcai.org/proceedings/2026/0322.pdf","markdown":"https://ijcai.org/proceedings/2026/322.md","bibtex":"https://ijcai.org/proceedings/2026/bibtex/322","volume":{"slug":"2026","title":"Proceedings of the Thirty-Fifth International Joint Conference on Artificial Intelligence","short_title":"IJCAI 2026","url":"https://ijcai.org/proceedings/2026/","year":2026,"location":"Bremen, Germany","dates":"15-21 August 2026","editors":"Diego Calvanese","isbn":"978-1-956792-09-6","publisher":"International Joint Conferences on Artificial Intelligence Organization","doi_prefix":"10.24963/ijcai.2026"},"section":"Data Mining","track":"Main Track","number":322,"pages":{"first":2897,"last":2905},"keywords":["Data Mining: Anomaly/outlier detection","Data Mining: Mining graphs"],"videos":[],"references":[],"license_note":"Copyright International Joint Conferences on Artificial Intelligence."}