Intel "big data" science and technology center vision and execution plan

Abstract

Intel has moved to a collaboration model with universities consisting of "Science and Technology Centers" (ISTCs). These are located at a "hub" university with participation from other universities, contain embedded Intel personnel, and are focused on some research theme. Intel held a national competition for a 5th Science and Technology center in 2012 and selected a proposal from M.I.T. with a theme of "Big Data". This paper presents the big data vision of this technology center and the execution plan for the first few years.

Data Science: Nature and Pitfalls

“…There are two ways to explore major research challenges: one is to summarize what concerns the relevant communities, and the other is to scrutinize the potential issues arising from the intrinsic complexities and nature of data science problems as complex systems [8], [8]. Taking the first approach, we can obtain a picture of the main research challenges by summarizing the main topics and issues in the statistics communities [16], [67], [63], informatics and computing communities [52], [11], vendors [56], government initiatives [61], [60], [19], [20], [59] and research institutions [62], [47], which focus on data science and analytics. The second approach is much more challenging, as it requires us to explore the unknown space of the complexities and comprehensive intelligence in complex data problems.…”

Big data: Evaluation criteria for big data analytics technologies

“…Big data is characterised by the following attributes: volume (Dong & Srivastava 2013;Letouze 2012;Liu 2013;McGuire Manyika & Michael 2012;Singh & Singh 2012), velocity (Agrawal et al 2009;Madden 2012;Stonebraker, Madden & Dubey 2013), variety (Letouze 2012), veracity (Dong & Srivastava 2013;Yan 2013) and value (Yan 2013). The definition of big data adopted in this article is derived from observations made by Yan (2013) on the concept of big data.…”

Materialization optimizations for feature selection workloads

“…Since then, there has been a recent resurgence in arrays as first-class citizens [14,15,31]. For example, Stonebraker et al [31] recently envisioned the idea of using carefully optimized C++ code, e.g., ScaLAPACK, in array databases for matrix calculations. Our Columbus system is complementary to these efforts, as we focus on how to optimize the execution of multiple operations to facilitate reuse.…”