Category: Web Technologies
Copyright © 2015, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited.
7584
Advanced Analytics for Big Data
INTRODUCTION
Big Data has emerged as one of the most challenging
aspects of business data and scientific processing within
the past 20 years. A major problem is that although we
may be able to collect the data, we often have inadequate
means for analyzing it. Kaisler, Armour, Espinosa,
and Money (2013) identified some of the issues and
challenges associated with using Big Data. INFORMS
defines analytics as “the process of transforming data
into insight for the purpose of making decisions”
(2013). It involves formulating specific problems or
questions; identifying, gathering and organizing the
relevant data; and selecting and applying the appropri-
ate methods, algorithms, heuristics and procedures to
solve the problems or answer the questions. Analytics
are quantitative and qualitative, linear and non-linear,
large and small, numerical versus symbolic, and vary
along other dimensions as well.
BACKGROUND
“Big Data” originally meant the volume of data that
could not be processed (efficiently) by traditional
database methods and tools. The original definition
focused on structured data, but most researchers and
practitioners realize that most of the world’s information
resides in unstructured data, largely in the form of text
and imagery, both still and video, and in audio. Today,
big data refers to data volumes in the range of tens of
petabytes (10
16
) and beyond. Such volumes exceed
the capacity of current on-line storage and processing
systems. Big Data was originally described by the 3Vs
(Laney 2001), but Kaisler, Armour, Espinosa, and
Money (2013) have suggest two more.
Up until 30 years ago, simple business models
often sufficed for international business. Globaliza-
tion, brought on by advances in digital technology,
information available at our fingertips, and a rapidly
changing, even chaotic, international political environ-
ment, has up-ended these models. It has increased the
diversity and uncertainty in outcomes when complex
systems such as financial flows and markets, regional
economies and political systems, and transnational
threats involving multiple actors are in constant flux.
Traditional big data analysis has focused on data
mining. The explosion of social media has provided
massive amounts of data that can be analyzed and ex-
ploited for a wide variety of purposes. Understanding
what patterns exist and determining how to use them
has been the primary problem for the past decade.
Identifying the relevant information from the plethora
of ambiguous and often contradictory data will require
sophisticated diagnostic, predictive, and prescriptive
analytical methods.
Advanced analytics is the application of multiple
analytic methods that address the diversity of big
data to provide descriptive results and to yield action-
able predictive and prescriptive results that facilitate
decision-making. Advanced analytics go beyond data
mining and statistical processing methods to encompass
logic-based methods, qualitative analytics, and non-
statistical quantitative methods. Advanced analytics
Stephen Kaisler
SHK and Associates, USA & George Washington University, USA
J. Alberto Espinosa
Kogod School of Business, American University, USA
Frank Armour
Kogod School of Business, American University, USA
William Money
School of Business Administration, George Washington University, USA
DOI: 10.4018/978-1-4666-5888-2.ch747