Analysis of data by humans can be a time-consuming activity and thus use of sophisticated cognitive systems can be utilized to crunch this enormous amount of data. Cognitive computing can be utilized to reduce the shortcomings of the concerns faced during big data analytics. The aim of the study is to provide readers a complete understanding of past, present and future directions in the domain big data and cognitive computing. A systematic literature review has been adopted for this study by using the Scopus, DBLP and Web of Science databases. The work done in the field of big data and cognitive computing is currently at the nascent stage and this is evident from the publication record. The characteristics of cognitive computing, namely observation, interpretation, evaluation and decision were mapped to the five V’s of big data namely volume, variety, veracity, velocity and value. Perspectives which touch all these parameters are yet to be widely explored in existing literature.

Introduction

Advancement in technology has always been a natural phenomenon. Drastic technological advancements often tend to miss out to find itself a mass market where it can be sold. The population of the world stands at more than 7.4 billion people (Worldometers, 2016) and more than 3.1 billion of these people are connected to the internet (eMarketer Report, 2015). More than 4.4 billion people are using mobile phone, out of this 1.86 billion (42%) of people are using smart phone (eMarketer Report, 2015). The number of devices and connections will keep increasing year on year basis and thus this had led to an explosion of data (Yaqoob et al., 2016). This enormous amount of data that is produced on a continuous basis is termed as big data (Kreps & Kimppa, 2015). Fosso-Wamba et al. (2015) have discussed the concept of big data in detail and have brought out the various definitions by other researchers in this domain. Big data analytics has gained significant amount of importance as it enables organizations to be ahead of their competitors (Li, Tao, Cheng, & Zhao, 2015; Habib ur Rehman, Chang, Batool, & Wah, 2016; Bumblauskas, Nold, Bumblauskas, & Igou, 2017). Companies want to analyze this raw data as they want to spot the trends that they can use for further profit maximization (Frizzo-Barker, ChowWhite, Mozafari, & Ha, 2016; Sun, Strang, & Firmin, 2016). The dire need to address the concerns of data deluge has led to the emergence of big data analytics (Dubey et al., 2015; Liberatore, Johnson, & Clain, 2016). Analysis of data by humans can be a time-consuming activity and thus use of sophisticated cognitive systems can be utilized to crunch this enormous amount of data (Kim, Chan, & Gupta, 2016). Intel CEO Brian Krzanich explains in his editorial article on artificial intelligence (AI) in 2016 that “AI is based on the ability of machines to sense, reason, act and adapt based on learned experience” (Krzanich, 2016). In a system that is based on AI, it works on the rules and parameters that are fed inside it whereas a cognitive computing based system works by intercepting the command and then drawing inferences and suggesting possible solutions. Cognitive computing is an AI based system that enables it to interact with humans like a fellow human, interpret the contextual meaning, analyze the past record of the user and draw deductions based on that interactive session. Cognitive computing helps the humans in decision making whereas AI based systems works on the concept that machines are capable of making better decisions on the human’s behalf. Cognitive computing is a sub-set of AI and anything that is cognitive is also AI. The goal of cognitive computing is to build a rational, combined and collective mechanism motivated by the capability of human mind (Kwon, Lee, & Shin, 2014).