Trivial Graph Features and Classical Learning are Enough to Detect Random Anomalies

2026-03-02

AI Summary

This study addresses the challenge of efficiently detecting randomly injected anomalous interactions in link streams when ground-truth labeled data are unavailable. The authors propose a lightweight detection approach that combines simple graph-based features—such as node degree and inter-event time intervals—with classical machine learning algorithms, including logistic regression and decision trees. The method achieves strong detection performance with minimal computational overhead and offers high interpretability. These results not only demonstrate the effectiveness of basic structural and temporal features coupled with traditional models for identifying random anomalies but also challenge the prevailing paradigm that relies on complex, resource-intensive architectures. The work thus opens new avenues for exploring more sophisticated anomaly types using similarly efficient and interpretable frameworks.

Technology Category

• Data Mining & Knowledge Management: Anomaly/Outlier Detection
• Machine Learning: Graph-based Machine Learning
• Reasoning under Uncertainty: Relational Probabilistic Models

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods
• Security and Privacy: Large-scale security measurements

Abstract

Detecting anomalies in link streams that represent various kinds of interactions is an important research topic with crucial applications. Because of the lack of ground truth data, proposed methods are mostly evaluated through their ability to detect randomly injected links. In contrast with most proposed methods, that rely on complex approaches raising computational and/or interpretability issues, we show here that trivial graph features and classical learning techniques are sufficient to detect such anomalies extremely well. This basic approach has very low computational costs and it leads to easily interpretable results. It also has many other desirable properties that we study through an extensive set of experiments. We conclude that detection methods should now target more complex kinds of anomalies.

Problem

• anomaly detection
• link streams
• random anomalies
• graph features
• classical learning

Innovation

• trivial graph features
• classical machine learning
• anomaly detection
• link streams
• interpretability