Page 116 - Building Big Data Applications

P. 116

Visualization, storyboarding and

applications

Most people use statistics the way a drunkard uses a lamp post, more for support
than illumination
Mark Twain

Visualization and dashboarding are very essential traits that every enterprise whether big
or small needs to use. The storyboards that can be painted with data are vast and can be
very simple to extremely complex. We have always struggled to deliver the business
insights; often business users have built their own subsystems and accomplished what
they needed. But the issue is we cannot deliver all the data needed for the visualization
and the dashboarding. As we progressed through the years, we have added more
foundations to this visualization by adding neural networks, machine learning, and
artiﬁcial intelligence algorithms. The question to answer is that the underlying data
engineering has progressed and become very useable, how to harness this power into the
storyboard and create the visual magic? Storyboarding is a very interesting way to
describe the visualization, reporting, and analytics we do in the enterprise. We create
different iterations of stories and it is consumed by different teams. How do we align the
story to the requirements? What drives the insights? Who will determine the granularity
of data and the aggregation requirements? What teams are needed to determine the
layers of calculations and the infrastructure required?
The focus of this chapter is to discuss how to build big data applications using the
foundations of visualization and storyboarding. How do we leverage and develop the
storyboard with an integration of new technologies, combine them with existing data-
bases and analytical systems, create powerful insights and on-demand analytical
dashboards that will deliver immense value? We will discuss several use cases of data
analytics and talk about data ingestion, especially large data sets, streaming data sets,
data computations, distributed data processing, replications, stream versus batch ana-
lytics, analytic formulas, once versus repetitive executions of algorithms, and supervised
and unsupervised learning and execution. We will touch speciﬁc areas on applications
around call center, customers, claims, fraud, and money laundering. We will discuss
how to implement robotic process automation, hidden layers of neural networks, and
artiﬁcial intelligence to deliver faster visualizations, analytics, and applications.
Let us begin the journey into the world of visualization and analytics like how we saw
the other chapters discuss vast data, complexity, integration, and rhythms of

111 112 113 114 115 116 117 118 119 120 121