Real-Time Analytics In Streaming Big Data: Techniques And Applications

Abstract

The increasing prevalence of streaming big data has revolutionized how organizations approach real-time analytics, providing a competitive edge by enabling immediate, actionable insights from continuously generated data streams. This review systematically examines real-time analytics techniques and their applications in streaming big data using the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) framework. The methodology involves identifying, screening, and synthesizing relevant studies to provide a comprehensive overview of state-of-the-art techniques, including data preprocessing, stream processing engines, distributed computing architectures, and machine learning algorithms specifically designed for high-velocity data streams. The review further categorizes and evaluates the applications of these techniques across key industries such as healthcare, financial services, e-commerce, and intelligent transportation systems. The results underscore the critical role of stream processing engines like Apache Kafka, Apache Flink, and Spark Streaming in managing data velocity and volume, while highlighting the growing importance of machine learning models in extracting real-time insights. Challenges such as scalability, fault tolerance, and latency issues are discussed, along with emerging solutions like edge computing and federated learning. The findings contribute to the evolving field of streaming big data analytics by providing insights into best practices and identifying research gaps for future exploration.