Volume 18, No. 6, 2021
Iot Data Preprocessing - A Survey
V.A. Jane , Dr. L. Arockiam
Abstract
The Internet of Things (IoT) is a rapidly evolving system in engineering and science. The sensors utilized in numerous sectors output a significant quantity of data. As a result, several consumers have a clear desire for efficient knowledge from these vast databases. This enormous dataset is far from flawless; it has several flaws (like distortion, inconsistent data, and anomalies) and is unsuitable for investigation due to the risk of inaccurate results. As a result, data preprocessing is a necessary approach for these information. Data preprocessing is a crucial and necessary phase, with the primary purpose of using procedures to filter, refine, repair, and enhance the raw information. The purpose of this study is to conduct a survey of IoT data preprocessing and methodologies. This article analyses current data preprocessing studies in the IoT environment, as well as the history of IoT data preprocessing and review articles of sophisticated data preprocessing approaches. The image clearly depicts the categorization of different preprocessing methodologies and procedures. Preprocessing cleaning, conversion, minimization, and integration methods are discussed. Furthermore, strategies for implementing such ideas in IoT data preprocessing are presented. IoT approaches for data preparation in diverse applications are listed. Lastly, difficulties and obstacles are explored that will be valuable in future research.
Pages: 2070-2080
Keywords: IoT, Preprocessing, Data Cleaning, Noise handling