A survey on big data classification

Abstract

Big data refers to vast volumes of structured and unstructured data that are too large or complex for traditional data-processing methods to handle efficiently. The importance of big data lies in its ability to provide actionable insights and drive decision-making across various industries, such as healthcare, finance, marketing, and government, by enabling more accurate predictions, and personalized services. Moreover, traditional big data classification approaches, often struggle with big data's complexity. They failed to manage high-dimensionality, deal with non-linearity, or process data in real time. For effective big data classification, robust computing infrastructure, scalable storage solutions, and advanced algorithms are required. This survey provides a thorough assessment of 50 research papers based on big data classification, by identifying the struggle faced by current big data classification techniques to process and classify data efficiently without substantial computational resources. The analysis is enabled on a variety of scenarios and key points. In this case, this survey will enable the classification of the techniques utilized for big data classification that is made based on the rule-based, deep learning-based, optimization-based, machine learning-based techniques and so on. Furthermore, the classification of techniques, tools used, published year, used software tool, and performance metrics are contemplated for the analysis in big data classification. At last, the research gaps and technical problems of the techniques in a way that makes the motivations for creating an efficient model of enabling big data classification optimal.