期刊名称：Bulletin of the Technical Committee on Data Engineering
出版年度：2017
卷号：40
期号：3
页码：4
出版社：IEEE Computer Society
摘要：From large-scale physical simulations, to high-throughput genomic sequencing, and from conversational agentinteractions, to sensor data from the Internet of Things, the need for data analytics—extracting insights fromlarge datasets—has never been greater. At the same time, current data analytics tools are powerless in harnessingthe hidden potential within these datasets. The bottleneck is not one of “scale”—we already know how to processlarge volumes of data quickly—but instead stems from the humans-in-the-loop. As dataset sizes have grown, thetime for human analysis, the cognitive load taken on by humans, and the human skills to extract value from data,have either stayed constant, or haven’t grown at a commensurate rate. Thus, at present, there is a severe lack ofpowerful tools that incorporate humans as a “first-class citizen” in data analytics, helping them interactivelymanage, analyze, and make sense of their large datasets.
