Investigation of Machine Learning Analytic-Based Data Classification Frameworks on Iot-Based Infrastructures

Abstract

The Internet of Things revolutionizes and enables convenient, automated lifestyles for people today. Over the past ten years, advances in computer, connectivity, and application design have led to its development. The Internet of Things has quickly spread around the world, impacting every person on the planet. Everyday IoT devices, such as smartphones, Google Home assistants, smart vehicles, and automated systems in buildings, including smart elevators and temperature control, as well as drones for environmental monitoring and leisure, play crucial roles in our daily lives. Therefore, the purpose of this research is to explore data classification frameworks based on machine learning analytics in Internet of Things-based infrastructures. This study focuses on articles published between 2010 and 2023, utilizing a comprehensive literature review approach. The study entails reading through at least eighty peer-reviewed conference proceedings, industry white papers, and journal articles. An overview of the relationship between machine learning and the Internet of Things (IoT) is given in this study. The approach consists of four standard processes that are used in review papers: gathering articles from Thomson Reuters Web of Science, identifying categories, carrying out descriptive analysis, and assessing the information acquired. A discussion on the categorization of diverse analytics methods for IoT is presented in the paper. Moreover, it introduces and discusses the architectural Framework for IoT and the challenges in IoT Data Classification. Furthermore, Machine Learning Strategies for Identifying and Classifying IoT were presented. Researchers must investigate machine learning analytic-based data classification frameworks for Internet of Things infrastructures. The challenges arising from the unique characteristics of IoT data call for innovative solutions, and current frameworks demonstrate promising results.