Unit 2

Using Big Data to Get Results, Basic analytics, Advanced

analytics, Operationalized analytics, Monetizing analytics,
Modifying Business Intelligence Products to Handle Big Data,
Analytical algorithms, Infrastructure support, Studying Big
Data Analytics Examples, Orbitz, Nokia, NASA, Big Data
Analytics Solutions.
Defining Big Data Analytics:
• Big data is most useful if you can do something with it, the question
becomes, how do you analyze it?
• Companies like Amazon and Google are masters at analyzing big data.
• And they use the resulting knowledge to gain a competitive advantage.
• Just think about Amazon’s recommendation engine.
• The company takes all your buying history together with what it
knows about you, your buying patterns, and the buying patterns of
people like you to come up with some pretty good suggestions.
• It is a marketing machine, and its big data analytics capabilities have
made it extremely successful.
• The capability to analyze big data provides unique opportunities for
your organization as well.
• Big data analytics has a wide range of applications, from fraud
detection and risk management to customer segmentation and
personalized marketing.
Using Big Data to get Results:
• The first question that you need to ask yourself before you dive into
big data analysis is what problem are you trying to solve? You may
not even be sure of what you are looking for.
• You know you have lots of data that you think you can get valuable
insight from. And certainly, patterns can emerge from that data before
you understand why they are there.
• For instance, are you interested in predicting customer behavior to
prevent churn(the rate at which customers stop doing business with a
company over a given period of time)? Do you want to analyze the
driving patterns of your customers for insurance premium purposes?
Are you interested in looking at your system log data to ultimately
predict when problems might occur?
• The kind of high-level problem is going to drive the analytics you
decide to use.
• Alternately, if you’re not exactly sure of the business problem you’re
trying to solve, maybe you need to look at areas in your business that
need improvement. Even an analytics-driven strategy — targeted at
the right area — can provide useful results with big data.
• When it comes to analytics, you might consider a range of
possible kinds, which are outlined in Table 12-1.
Basic analytics
• Basic analytics can be used to explore your data, if you’re not sure what you
have, but you think something is of value.
• This might include simple visualizations or simple statistics.
• Basic analysis is often used when you have large amounts of disparate data.
Here are some examples:
Slicing and dicing:
• Slicing and dicing refers to breaking down your data into smaller sets of
data that are easier to explore.
• For example, you might have a scientific data set of water column data from
many different locations that contains numerous variables captured from
multiple sensors.
• Attributes might include temperature, pressure, transparency, dissolved
oxygen, pH, salinity, and so on, collected over time.
• You might want some simple graphs or plots that let you explore your data
across different dimensions, such as temperature versus pH or transparency
versus salinity.
• The point is that you might use this basic type of exploration of the
variables to ask specific questions in your problem space.
Basic monitoring:
• You might also want to monitor large volumes of data in
real time.
• For example, you might want to monitor the water
column attributes in the preceding example every second
for an extended period of time from hundreds of locations
and at varying heights in the water column. This would
produce a huge data set.
• Or, you might be interested in monitoring the buzz
associated with your product every minute when you
launch an ad campaign.
• Whereas the water column data set might produce a large
amount of relatively structured time-sensitive data, the
social media campaign is going to produce large amounts
of disparate kinds of data from multiple sources across
the Internet.
Anomaly identification:
• You might want to identify anomalies, such as an event where
the actual observation differs from what you expected, in your
data because that may clue you in that something is going
wrong with your organization, manufacturing process, and so
on.
• For example, you might want to analyze the records for your
manufacturing operation to determine whether one kind of
machine, or one operator, has a higher incidence of a certain
kind of problem.
• This might involve some simple statistics like moving
averages triggered by an alert from the problematic machine.
Advanced analytics
• Advanced analytics provides algorithms for complex analysis of
either structured or unstructured data.
• It includes sophisticated statistical models, machine learning,
neural networks, text analytics and other advanced data-mining
techniques.
• Among its many use cases, advanced analytics can be deployed to
find patterns in data, prediction, forecasting, and complex event
processing.
• Today, advanced analytics is becoming more mainstream.
• With increases in computational power, improved data
infrastructure, new algorithm development, and the need to obtain
better insight from increasingly vast amounts of data, companies
are pushing toward utilizing advanced analytics as part of their
decision-making process.
• Businesses realize that better insights can provide a superior
competitive position.
Here are a few examples of advanced analytics for big data:
Predictive modeling:
• Predictive modeling is one of the most popular big data advanced
analytics use cases.
• A predictive model is a statistical or data-mining solution
consisting of algorithms and techniques that can be used on both
structured and unstructured data (together or individually) to
determine future outcomes.
• For example, a telecommunications company might use a
predictive model to predict customers who might drop its service.
• In the big data world, you might have large numbers of predictive
attributes across huge amounts of observations.
• Whereas in the past, it might have taken hours (or longer) to run a
predictive model, with a large amount of data on your desktop,
you might be able to now run it iteratively hundreds of times if
you have a big data infrastructure in place.
Text analytics:
• Unstructured data is such a big part of big data, so text
analytics — the process of analyzing unstructured text,
extracting relevant information, and transforming it into
structured information that can then be leveraged in various
ways — has become an important component of the big data
ecosystem.
• The analysis and extraction processes used in text analytics
take advantage of techniques that originated in computational
linguistics, statistics, and other computer science disciplines.
• Text analytics is being used in all sorts of analysis, from
predicting churn, to fraud, and to social media analytics.
Other statistical and data-mining algorithms:
• This may include advanced forecasting, optimization, cluster
analysis for segmentation or even micro segmentation, or
affinity analysis(is used for studying patterns and behavior in
customers to determine links in purchases).

• Advanced analytics doesn’t require big data. However,

being able to apply advanced analytics with big data can
provide some important results.
What is data mining?
• Data mining involves exploring and analyzing large amounts
of data to find patterns in that data.
• The techniques came out of the fields of statistics and artificial
intelligence (AI), with a bit of database management thrown
into the mix.
• Generally, the goal of the data mining is either classification or
prediction. In classification, the idea is to sort data into groups.
• For example, a marketer might be interested in the
characteristics of those who responded versus who didn’t
respond to a promotion. These are two classes.
• In prediction, the idea is to predict the value of a continuous
(that is, nondiscrete) variable.
• For example, a marketer might be interested in predicting
those who will respond to a promotion.
Operationalizing analytics:
• It refers to the process of integrating analytical models, insights, and results
into an organization's day-to-day decision-making processes and systems.
• It involves bridging the gap between data science and business operations by
implementing analytics in a way that drives actionable outcomes and creates
value for the organization.
• When you operationalize analytics, you make them part of a business
process.
• For example, a model could be built to predict customers who are good
targets for upselling when they call into a call center.
• The call center agent, while on the phone with the customer, would receive a
message on specific additional products to sell to this customer.
• The agent might not even know that a predictive model was working behind
the scenes to make this recommendation.
• So, operationalizing analytics involves a combination of technical and
organizational changes.
• By integrating analytics into your organization's day-to-day decision-making
processes and systems, you can drive better outcomes, create value, and gain
a competitive advantage.
• Operationalized analytics involves a combination of
technologies, processes, and people.
• The process begins with the collection of data from various
sources, such as sensors, logs, and databases.
• This data is then processed and analyzed to extract insights
and identify patterns.
• The insights are then presented to decision-makers in a format
that is easy to understand and actionable.
• One of the key components of operationalized analytics is the
use of machine learning algorithms.
• Machine learning algorithms can analyze large datasets and
identify patterns that are not immediately obvious to humans.
• These algorithms can also learn from the data and improve
their accuracy over time.
• Operationalized analytics can be applied to a variety of industries,
including healthcare, finance, manufacturing, and transportation.
• For example, in healthcare, operationalized analytics can be used to
analyze patient data and identify trends that can be used to improve
patient outcomes.
• In finance, operationalized analytics can be used to detect
fraudulent transactions and improve risk management.
• To implement operationalized analytics, organizations need to have
the right technology infrastructure, data management processes, and
skilled staff.
• It requires a combination of technical expertise, domain knowledge,
and business acumen.
• With the right combination of technology, people, and processes,
operationalized analytics can provide significant benefits to
organizations, including improved efficiency, reduced costs, and
better decision-making.
Monetizing analytics:
• Monetizing analytics involves turning your organization's data and analytics
capabilities into revenue-generating products or services.
• This can be done in a variety of ways, depending on the specific needs and
goals of your organization.
• Here are some key strategies for monetizing analytics:
• Sell analytics as a service: One way to monetize analytics is to offer analytics
as a service to other organizations. This involves providing data analysis,
modeling, and insights on a subscription basis. This can be done through a
cloud-based platform, which enables customers to access analytics capabilities
remotely.
• Build and sell predictive models: Another strategy is to build predictive
models that can be sold to other organizations. This involves identifying
common business problems that can be solved using predictive analytics, such
as churn prediction or fraud detection, and developing models that can be
customized for different customers.
• Develop data products: Another way to monetize analytics is to develop data
products that can be sold to customers. This involves identifying data sets that
are valuable to other organizations, and packaging them in a way that is easy to
use and access.
• Create value-added services: Analytics can also be monetized
by creating value-added services that leverage analytics
insights. For example, an organization that provides supply
chain management services might use analytics to optimize
inventory levels, and then offer this as a value-added service to
its customers.
• Create partnerships and collaborations: Finally, partnerships
and collaborations can be a powerful way to monetize analytics.
This involves teaming up with other organizations to offer joint
products or services that leverage each other's analytics
capabilities.
• Hence, monetizing analytics requires a deep understanding of
your organization's data and analytics capabilities, as well as the
needs and goals of your customers. By leveraging analytics to
create new products and services, you can not only generate
new revenue streams, but also differentiate yourself in a
crowded marketplace and gain a competitive advantage.
Modifying Business Intelligence Products to Handle Big
Data:
• Traditional business intelligence products weren’t really
designed to handle big data.
• They were designed to work with highly structured, well-
understood data, often stored in a relational data repository
and displayed on your desktop or laptop computer.
• This traditional business intelligence analysis is typically
applied to snapshots of data rather than the entire amount of
data available.
• As we discuss in previous chapter, big data consists of
structured, semi-structured, and unstructured data.
• You often have a lot of it, and it can be quite complex.
• When you think about analyzing it, you need to be aware of
the potential characteristics of your data:
It can come from untrusted sources.
• Big data analysis often involves aggregating data from various sources.
• These may include both internal and external data sources. How
trustworthy are these external sources of information?
• For example, how trustworthy is social media data like a tweet? The
information may be coming from an unverified source.
• The integrity of this data needs to be considered in the analysis.
It can be dirty.
• Dirty data refers to inaccurate, incomplete, or erroneous data.
• This may include the misspelling of words; a sensor that is broken, not
properly calibrated, or corrupted in some way; or even duplicated data.
• Data scientists debate about where to clean the data — either close to the
source or in real time.
• Of course, one thought says that the dirty data should not be cleaned at all
because it may contain interesting outliers.
• The cleansing strategy will probably depend on the source and type of data
and the goal of your analysis.
• For example, if you’re developing a spam filter, the goal is to detect the
bad elements in the data, so you would not want to clean it.
The signal-to-noise ratio can be low.
• In other words, the signal (usable information) may only be a tiny
percent of the data; the noise is the rest.
• Being able to extract a tiny signal from noisy data is part of the
benefit of big data analytics, but you need to be aware that the signal
may indeed be small.
Analytical algorithms
• When you’re considering big data analytics, you need to be
aware that when you expand beyond the desktop, the
algorithms you use often need to be refactored, changing the
internal code without affecting its external functioning.
• The beauty of a big data infrastructure is that you can run a
model that used to take hours or days in minutes.
• This lets you iterate on the model hundreds of times over.
Some common analytical algorithms include:
• Regression analysis: Used to identify the relationship between
two or more variables and predict future values based on
historical data.
• Decision trees: Used to classify data by creating a tree-like
model of decisions and their possible consequences.
• Clustering: Used to group similar data points together based on
their characteristics.
• Principal component analysis: Used to identify the most
important factors or components in a dataset.
• Association rules: Used to find relationships between different
items in a dataset and identify patterns.
• Neural networks: Used to simulate the behavior of the human
brain to analyze and process complex data.
• Random forests: Used to build a model of decision trees and
combine the results to make predictions.
• Analytical algorithms can be used to solve a wide range of
problems, including predicting customer behavior, identifying
fraud, optimizing production processes, and diagnosing
medical conditions.
• The selection of the appropriate analytical algorithm depends
on the specific problem at hand and the type of data being
analyzed.
• Different algorithms may have different strengths and
weaknesses, and selecting the right algorithm can have a
significant impact on the accuracy and usefulness of the
insights extracted from the data.
Infrastructure support
To support big data, you need a robust infrastructure that can
handle the volume, velocity, and variety of data. Here are some
key components of the infrastructure needed to support big data:
Integrate technologies:
• The infrastructure needs to integrate new big data technologies
with traditional technologies to be able to process all kinds of
big data and make it consumable by traditional analytics.
Store large amounts of disparate data:
• A Hadoop system may be needed that can process/store/manage
large amounts of data at rest, whether it is structured, semi-
structured, or unstructured.
Process data in motion:
• A stream-computing capability may be needed to process data in
motion that is continuously generated by sensors, smart devices,
video, audio, and logs to support real-time decision making.
Warehouse data:
• To support analytics and reporting on big data, a data warehousing
solution is required. Technologies like Apache Hive, provide data
warehousing capabilities for big data.
Data Visualization and Reporting:
• Big data infrastructure should support tools for data visualization and
reporting to make sense of the data. Popular tools include Tableau,
Power BI, and Apache Superset, which provide interactive dashboards
and visualizations for data analysis.
Scalability and Fault Tolerance:
• Big data infrastructure needs to be highly scalable and fault-tolerant to
handle the ever-increasing volume of data. Technologies like Apache
Hadoop, distributed databases, and cloud computing platforms like
Amazon Web Services (AWS) or Google Cloud Platform (GCP)
provide scalability and fault tolerance features.

These are some of the essential components of infrastructure needed to

support big data.
Studying Big Data Analytics Examples
Orbitz:
• Orbitz is a well-known online travel agency that provides
customers with a platform to book flights, hotels, car
rentals, and vacation packages.
• The company was established in 1999, and its website
went live in 2001.
• Users of Orbitz perform over a million searches a day,
and the company collects hundreds of gigabytes of raw
data each day from these searches.
• Big data analytics plays a crucial role in Orbitz's
operations, helping the company analyze vast amounts of
data to improve its services and enhance customer
experience.
Some examples of how Orbitz utilizes big data analytics:
• Personalized recommendations: Orbitz leverages big data
analytics to analyze customer data, including past
bookings, preferences, and browsing behavior, to provide
personalized travel recommendations. By analyzing this
data, Orbitz can suggest tailored flight and hotel options,
vacation packages, and even personalized discounts and
offers based on individual customer profiles.
• Price optimization: Big data analytics enables Orbitz
to monitor and analyze massive amounts of real-time
data on flight prices, hotel rates, and other travel-related
costs. By leveraging historical data, market trends,
and competitor analysis, Orbitz can optimize its
pricing strategies to offer competitive rates, discounts,
and promotions, while still ensuring profitability.
• Fraud detection and prevention: With the rise in online
fraud, Orbitz relies on big data analytics to identify and
prevent fraudulent activities. By analyzing vast amounts of
customer data, including booking patterns, payment
information, and user behavior, Orbitz can detect
anomalies and potential fraud instances. This helps the
company safeguard its customers' financial information and
maintain a secure booking environment.
• Demand forecasting: Big data analytics allows Orbitz to
predict and forecast demand for flights, hotels, and other
travel services. By analyzing historical booking patterns,
seasonal trends, and external factors such as events and
holidays, Orbitz can optimize inventory management,
pricing, and availability. This helps ensure that customers
have access to the services they need and enables Orbitz to
maximize revenue and customer satisfaction.
• Customer sentiment analysis: Orbitz utilizes big data
analytics to analyze customer feedback and sentiment across
various channels, including reviews, social media, and
customer support interactions. By analyzing this data,
Orbitz can gain insights into customer satisfaction, identify
areas for improvement, and promptly address any concerns or
issues. This helps Orbitz enhance its overall customer
experience and build long-term customer relationships.
• These examples illustrate how Orbitz leverages big data
analytics to improve its operations, enhance customer
satisfaction, and stay competitive in the online travel
industry. By harnessing the power of data, Orbitz can
make data-driven decisions, provide personalized
recommendations, optimize pricing, detect fraud,
forecast demand, and analyze customer sentiment
effectively.
Nokia:
• Nokia, a global telecommunications company, utilizes big
data analytics in various aspects of its operations to
improve efficiency, enhance customer experience, and
drive innovation.
• Here are a few examples of how Nokia leverages big data
analytics:
• Network optimization: Nokia employs big data analytics
to analyze massive amounts of network data, including
network performance metrics, user behavior, and
traffic patterns. By leveraging this data, Nokia can
optimize its network infrastructure, identify and resolve
network congestion issues, improve service quality, and
proactively address network-related challenges.
• Customer experience management: Nokia utilizes big data
analytics to gain insights into customer behavior, preferences,
and satisfaction levels. By analyzing customer data from various
touchpoints, such as call records, network logs, and customer
feedback, Nokia can identify pain points, detect potential service
issues, and make data-driven decisions to improve customer
experience. This allows Nokia to deliver tailored solutions,
personalized offerings, and proactive customer support.
• Supply chain optimization: Big data analytics plays a
crucial role in Nokia's supply chain management. By
analyzing supply chain data, including inventory levels,
production schedules, supplier performance, and market
demand, Nokia can optimize its procurement, production,
and distribution processes. This helps minimize costs,
reduce lead times, improve inventory management, and
enhance overall supply chain efficiency.
• Product development and innovation: Nokia leverages big
data analytics to gain insights into market trends, customer
preferences, and emerging technologies. By analyzing large
volumes of data from various sources, including social
media, customer surveys, and market research, Nokia can
identify new product opportunities, validate ideas, and
develop innovative solutions to meet evolving customer
needs. This data-driven approach helps Nokia stay
competitive in a rapidly changing industry.
• These examples highlight how Nokia utilizes big data
analytics to drive operational efficiency, improve
customer experience, optimize supply chain management,
and foster innovation. By harnessing the power of data,
Nokia can make informed decisions, enhance network
performance, deliver personalized solutions, and stay at
the forefront of the telecommunications industry.
NASA
• NASA (National Aeronautics and Space Administration)
extensively utilizes big data analytics to support various
missions, research projects, and scientific endeavors. Here
are a few examples of how NASA leverages big data
analytics:
• Earth observation and climate analysis: NASA collects
and analyzes vast amounts of data from satellites, sensors,
and other sources to study Earth's climate, weather
patterns, and environmental changes.
– Big data analytics allows NASA scientists to process and
analyze this data to understand climate dynamics, track changes
over time, and develop models for climate prediction. This
information is crucial for monitoring and managing the Earth's
environment and informing policies related to climate change.
• Space exploration and mission planning: NASA relies on big
data analytics to analyze data from space missions, including data
from telescopes, rovers, and spacecraft.
– By processing and interpreting this data, NASA can discover new
celestial objects, explore distant planets, study the universe's origins, and
plan future missions. Big data analytics helps NASA scientists uncover
patterns, relationships, and anomalies in the vast amount of space data,
enabling groundbreaking discoveries and advancements in space
exploration.
• Spacecraft telemetry analysis: NASA uses big data analytics to
analyze telemetry data received from spacecraft and satellites.
This data includes information on spacecraft health, performance,
and scientific measurements. By analyzing this telemetry data in
real-time, NASA can monitor the status of its missions, detect
anomalies or malfunctions, and take appropriate actions to ensure
mission success.
– Big data analytics plays a crucial role in spacecraft operations, enabling
NASA to make data-driven decisions and optimize mission performance.
• Planetary exploration and rover data analysis: NASA's
missions to other planets, such as Mars rovers like Curiosity
and Perseverance, generate large volumes of data. Big data
analytics allows NASA to analyze the data collected by these
rovers, including images, spectra, and environmental
measurements. By leveraging big data techniques, NASA can
study planetary geology, search for signs of life, and make
discoveries about the Martian environment. Big data analytics
assists in extracting meaningful information from complex
datasets, enabling scientists to derive valuable insights from
planetary missions.
• These examples demonstrate how NASA utilizes big data
analytics to support a wide range of scientific research, space
exploration, and climate analysis. By harnessing the power of
data, NASA can advance our understanding of the universe,
monitor and protect our planet, and push the boundaries of
space exploration.