Classroom Companion: Business

20.1 · Zettabyte Era

---

312
20
20.2
 Characteristics of Big Data
Big data are defined by the following characteristics (see 
.
Fig. 
20.4
) referred to as 
the four Vs:
5
Volume, referring to the amount of digital data available for analysis.
5
Variety, referring to the richness of data categories. During the Zettabyte Era, 
stored data has evolved from being mainly structured data—data that is orga-
nized and adheres to a specific format—to unstructured data such as fusion of 
text, location data, video, images, and social media activity. Such complex data 
structures cannot be processed or analyzed by simple analytical tools.
5
Velocity, referring to the speed by which data is generated and processed. Big 
data is usually produced in continuous processes. Some of this data is captured, 
processed, and published in real time.
Data volume
Data velocity
Data veracity
Data variety
MB
GB
TB
PB
EB
ZB
Real time
(ms)
Periodic
(s)
Batch
(minutes)
Offline
(days)
Table
Database
Multi-
media
Social
media
Verified
Local
Non-verfied
Distributed
Multi-
provider
. Fig. 20.4 Growth in key characteristics of big data. (Authors’ own figure)
 
Chapter 20 · Big Data Economics

---

313
20
5
Veracity, referring to the exactness of the data. The accuracy of the data analy-
sis depends on biases, noise, inaccuracies, and irregularities in the data set. 
Insofar as possible, the data analysis should take such anomalies into account 
and estimate the validity of the result of the analysis.
The huge volumes of data are generated by both people and machines. Data pro-
duced by people include photos, text, videos, movies, music, professional content, 
video conferencing, and chat logs. Data produced by machines (M2M) include 
sensor data (e.g., measurement of environmental data, health monitoring, and 
intelligent traffic management), medical images (in particular MRI), videos from 
surveillance cameras, and system updates and configurations. The annual com-
pound growth rates for the various segments of M2M communication are expected 
to be between 10% and 50% from 2018 to 2023 (Holst, 
2018
).
Big data analytics includes tools and techniques to convert vast amounts of raw 
data into meaningful information that can be used as tradable goods or for optimi-
zation or personalization of services offered to users and customers. The input to 
the big data analytics algorithms is the vast amount of digital data that is collected 
about users, processes, and events. The output is information used in marketing, 
business planning, behavioral control, trend analysis, and statistics. Unstructured 
data turned into meaningful information is the basis for the value propositions of 
the big data services.
An example of a computer facility designed for big data storage and processing 
is shown in 
7
Box 
20.2
.