Real-time Big Data Analytics
The volume, velocity, and variety of digital data is increasing day by day. It is expected that by 2020, the amount of digital data will reach 40 trillion gigabytes, which was merely 1.2 trillion gigabytes in 2010. Therefore, organizations are increasingly pivoting towards leveraging big data technologies (e.g., Hadoop, Spark, and Cassandra) to deal with the massive volume, velocity, and variety of data referred to as ‘big data'. Collecting, storing, analyzing, and visualizing big data in real-time is the need of the hour in several domains such as healthcare, cyber security, and traffic management. For instance, rapid response based on large-scale data analysis mitigates the repercussions of an emergency situation such as traffic accidents and natural disasters. Similarly, real-time detection of cyber-attacks mitigates the impact of the attack by 97%.
Within the area of big data analytics, CREST researchers leverage state-of-the-art techniques (e.g., AI and search-based optimization) to design, implement, deploy, and evaluate big data systems for optimally collecting, storing, analyzing, and visualizing a large volume of data in real-time. CREST research particularly focuses on the evaluation of big data storage solutions (e.g., Cassandra and MongDB) and big data analytical solutions (e.g., Spark and Flink) as deployed on private, public, and hybrid clouds. The application domains of our research on real-time big data analytics include but not limited to cyber security, oil and gas, and healthcare.