Classification of research efforts in dynamic/big data analytics

Abstract

The recent explosion in Dynamic (a.k.a., "Big") Data Analytics1 research provides a massive amount of software capabilities, published papers, and conference proceedings that make it difficult to sift through and inter-relate it all. This paper proposes a trial classification scheme with several orthogonal dimensions of classification. These dimensions include stages of application, challenges, solution origins, specialization of technologies, purpose, ownership (business type), data processing (batch vs. streaming), and data types applied to (structured, semi-structured and unstructured). The full list of determined categories in each dimension is presented. The classification scheme is intentionally made to be not too complex, to help anyone entering the expanding world of Big Data Analytics, by helping them gain a better understanding of the applicability of various tools and capabilities that are available, and how they contrast and synergize amongst each another. Additionally, this work can help with creation of educational materials, demarcation of the domain, and encourage full research coverage in big data analytics, as well as enable discovery and articulation of common principles and solutions. The research topics used in testing this classification scheme are retrieved from the top 20 most relevant papers of Scopus online database, which is aiming to be the largest repository of the peer-reviewed literature, as well as by reviewing examples of similar past classification attempts.