Location-Based Deep Collaborative Filtering for Quality of Service Prediction in IoT

In the realm of web services, various providers offer many analogous options. The selection of the optimal service from these alternatives relies heavily on Quality of Service (QoS), a metric encompassing non-functional attributes. To enhance QoS assessment for web services, a variety of QoS prediction methods have been proposed. Among these methodologies, collaborative web service QoS prediction emerges as one of the most prominent techniques. Recently, methods based on collaborative filtering (CF) have garnered substantial attention for service recommendations. However, traditional CF predominantly handles low-dimensional and linear interactions between users and services, often grappling with real-world data sparsity challenges. To address these limitations, this article introduces a novel deep CF model for service recommendation. The proposed model, inspired by deep learning and two similarity measures, solves the challenge of traditional collaborative filtering methods and provides a platform for fundamental improvements.