SD-MTMA: Integrating Static and Dynamic Methods for Detecting Multi-Trigger Multi-Action Rule Conflicts in Smart Homes

Smart homes are a key application of the Internet of Things, where users employ Trigger-Action Programming rules for personalized control. However, as the number of rules increases, conflicts arise, disrupting system operations. This study proposes a novel method to automatically detect these conflicts by combining static and dynamic approaches. The static method detects and resolves rule conflicts, while dynamic detection is performed across multiple scenarios generated from user-defined rules, enhancing accuracy. For conflicts involving device impacts, this study proposes the “opposing function extraction method,” and uses cosine similarity to calculate environmental impacts, minimizing manual intervention. A dataset of 20,537 complex rules was constructed for validation. Experimental results show that the method achieves a precision of 0.95, recall of 0.81, and F1-score of 0.88. Compared with UTEA, CDAM, and KGIBE, this approach significantly reduces manual annotation needs and improves conflict detection, CPU usage, and memory efficiency.