Anomaly detection plays a crucial role in ensuring the health and well-being of mothers by monitoring various health parameters in real-time. Unlike static health data, maternal health monitoring data are dynamic, possessing properties such as infiniteness, correlations, and changes in data distribution. These properties present challenges for current anomaly detection approaches. Firstly, saving the entire dataset is impractical due to its infiniteness. Secondly, existing methods often fail to consider correlations between different health metrics. Thirdly, changes in data distribution are not adequately addressed because of the absence of model update and change detection strategies. To tackle these issues, we propose Maternal health Anomaly Detection (MAD), a groundbreaking method combines sliding window mechanisms, model adjustments, and change detection strategies within Locality-Sensitive Hashing isolation Forest (LSHiForest) to enhance anomaly detection accuracy and efficiency while ensuring better scalability. This method, termed MAD, includes several key features: (1) The sliding window approach effectively manages the endless nature of data streams; (2) Incorporating Singular Value Decomposition (SVD) takes into account the interrelationships among different health metrics; (3) The change detection process, along with model updates, swiftly recognizes shifts in data distribution and retrains models as needed. Comprehensive experiments on relevant maternal health datasets validate the performance of MAD. Results indicate that MAD outperforms existing methods in terms of accuracy, efficiency, and scalability, making it a promising solution for maternal health monitoring.

In the Internet of Things (IoT) environment, user-service interaction data are often stored in multiple distributed platforms. In this situation, recommender systems need to integrate the distributed user-service interaction data across different platforms for making a comprehensive recommendation decision, during which user privacy is probably disclosed. Moreover, as user-service interaction records accumulate over time, they significantly reduce the efficiency of recommendations. To tackle these issues, we propose a lightweight and privacy-preserving service recommendation approach named SerRec_L2H. In SerRec_L2H, we employ Learning to Hash (L2H) to encapsulate sensitive user-service interaction data into less-sensitive user indices, which facilitates identifying users with similar preferences efficiently for accurate recommendations. We then validate the feasibility of our proposed SerRec_L2H approach through massive experiments conducted on the popular WS-DREAM dataset. The comparative analysis with other competitive approaches demonstrates that our proposal surpasses other approaches in terms ofrecommendation accuracy and efficiency while protecting user privacy.