A semantically enabled architecture for interoperable edge‐cloud continuum applied to the e‐health scenario

Angelo Martella,Antonella Longo,Marco Zappatore,Beniamino Di Martino,Antonio Esposito
DOI: https://doi.org/10.1002/spe.3375
2024-10-02
Software Practice and Experience
Abstract:The progress made in the field of medicine and the consequent increase in the prospect of life have contributed to rise people's interest towards a healthier lifestyle. Fitness activity is becoming a must for those who aspire to live more and better. However, this should be accompanied by additional good practices to safeguard individuals' life from risks that could undermine their health. Most of these risks are linked to personal, surrounding, and contextual conditions that technology can detect and monitor. Recommender systems can adequately support fitness activity by performing data analyzes aimed at identifying possible risk factors for users, starting from their physiological data and those related to the closest context where they are. This article introduces the architecture of a recommender system called App4Health in the context related to both mobile crowd sensing and wellness. The App4Health architecture consists of a smart application platform, capable of interfacing and managing data from heterogeneous edge sources, such as mobile phones, IoT, and sensors. The analysis result consists of the semantic generation of healthy behavioral conducts to the user as Telegram BOT messages. For evaluating the proposed solution, the article also provides a case study and a testbed. The testbed consists of a comparative stress test of two edge software components of the App4Health's architecture in order to identify the performance degradation threshold of these components, assuming that they can be deployed on edge‐level hardware devices with different technical specifications and configurations.
computer science, software engineering
What problem does this paper attempt to address?