A Unified XAI-NSAI Analytical Framework for Trustworthy IoT Anomaly Detection
Organizations: Department of Computer Science University of Cincinnati Cincinnati, OH, USA
Abstract
The widespread deployment of Internet of Things (IoT) devices in critical environments makes anomaly detection essential for identifying abnormal behaviors and supporting timely investigation and response. However, accurate detection alone is insufficient. Practical IoT deployments must also provide explanations, incorporate domain knowledge and reasoning, operate under IoT constraints, and support human decision making. Although recently proposed approaches increasingly incorporate Explainable Artificial Intelligence (XAI) or Neuro-Symbolic Artificial Intelligence (NSAI), XAI and NSAI methods are often considered independently. We argue that they should be considered together to better understand their complementary roles in supporting trustworthy IoT anomaly detection. To address this gap, we propose a unified analytical framework organized around five complementary dimensions. The framework provides a structured basis for characterizing existing approaches, identifying capability gaps and design trade-offs, and guiding the design of new anomaly detection systems for different IoT applications. We demonstrate the framework's analytical utility by applying it to three representative studies and illustrate its design utility through its application to a Medical IoT anomaly detection system.
Figures & tables
| Dimension | Key Question | Representative Scope |
| D1 Explanation Scope and Stage | What is explained, and when? | Intrinsic/ante-hoc; post-hoc; local/global; concept-level; temporal; hybrid |
| D2 Symbolic Reasoning Integration | How is symbolic knowledge and reasoning incorporated? | None; rule-based; knowledge-based; integration with learning, inference, or decision making |
| D3 Architectural Design | How are detection, explanation, and reasoning organized? | Learning-only; pipeline-based; hybrid; tightly integrated |
| D4 Deployment Context | Where are detection, explanation, and reasoning executed? | Centralized/cloud; edge; edge-cloud; federated/distributed |
| D5 User-Centered Explainability | How do explanations support intended users? | Technical; operational; domain-specific; role-, expertise-, task-, or feedback-aware |
| Study | D1 | D2 | D3 | D4 | D5 | Framework Interpretation |
| Rathi and Shandilya [ 26 ] | 2 | 0 | 2 | 3 | 2 | deployment-aware XAI pipeline, but no symbolic reasoning |
| Kalutharage et al. [ 20 ] | 3 | 2 | 2 | 3 | 2 | Explanation through cybersecurity knowledge/domain, but symbolic reasoning is largely inference-level |
| Bizzarri et al. [ 19 ] | 1 | 3 | 3 | 1 | 1 | deeper neural-symbolic integration, but no strong analyst facing explanations or deployment validation |
Interpretable Anomaly Detection
Interpretable anomaly detection identifies unusual data and explains the features, patterns, or conditions behind each alert.
See all 83 papers