Uber's Technology Stack and Data Analytics

This assignment delves into the technological foundation of Uber's operations. It examines their tech stack, highlighting key components like Hadoop and Spark, and explores how Uber leverages big data analytics to improve service efficiency, optimize rides, and enhance driver safety. The assignment draws upon various sources, including official Uber blog posts, technical articles, and industry reports.

Running Head: BUSINESS INTELLIGENCE USING BIG DATA – UBER CASE STUDY
Business intelligence using big data
Uber case study
Student’s name:
Institution Affiliation:

BUSINESS INTELLIGENCE USING BIG DATA
Table of Contents
Introduction to Uber...................................................................................................................................3
Use of Big data............................................................................................................................................3
Business strategy.........................................................................................................................................3
Improve user experience to drive more profit to drive more profitable customer engagements...........3
Business Initiative........................................................................................................................................4
Improve match results.............................................................................................................................4
Increase customers..................................................................................................................................4
Proper allocation of services using big data analytics..............................................................................4
Enhance security using driver Selfies powered by Microsoft Cognitive Services.....................................4
Outcomes and Critical Success Factors (CSF)...............................................................................................5
Optimize ratings and feedback system....................................................................................................5
Pricing value of the journey.....................................................................................................................5
Generate profile for passenger and driver using rating system...............................................................5
Allocate resources by analyzing public transport networks....................................................................6
Use business intelligence to develop algorithms for monitoring traffic data and journey time in real
time.........................................................................................................................................................6
Evaluate the match regarding proximity and fare...................................................................................6
Implement cashless system for payment integration..............................................................................6
Analyze and Implement market strategies like referrals, reviews, partnership, a loyalty.......................6
Data sources................................................................................................................................................7
Driver data...............................................................................................................................................7
Rider data................................................................................................................................................7
Payment data..........................................................................................................................................7
Traffic data...............................................................................................................................................7
GPS data..................................................................................................................................................7
Marketing data........................................................................................................................................8
Mobile data.............................................................................................................................................8
Uber Technology Stack................................................................................................................................8
HTTP Server Technologies.......................................................................................................................8
Client libraries..........................................................................................................................................8
Programming language............................................................................................................................8
Server Libraries........................................................................................................................................9
Databases................................................................................................................................................9

BUSINESS INTELLIGENCE USING BIG DATA
Data stores..............................................................................................................................................9
Server Software.......................................................................................................................................9
Operating System....................................................................................................................................9
Tasks..........................................................................................................................................................11
Role of socio media in decision-making.....................................................................................................11
Data Visualizing and Analytics...................................................................................................................11
MDM to support DS&BI.............................................................................................................................12
Support of NoSQL for Big Data Analytics...................................................................................................12
How Uber has used NoSQL Databases in Big Data.....................................................................................13
Big Data Value creation process................................................................................................................13
Conclusion.................................................................................................................................................14
References.................................................................................................................................................15

BUSINESS INTELLIGENCE USING BIG DATA
Introduction to Uber
Uber is a smartphone-based company with its headquarters in San Francisco. The
company is currently operating in six hundred and thirty-three cities all over the world. Uber
develops, markets and operates the Uber transport and food delivery apps. Uber uses drivers who
either have their cars, but also drivers have the option of renting out cars to the driver under their
umbrella. Uber uses a web portal uber.com and mobile applications –Android and iOS in its
systems.
Uber is based on big data principle of crowdsourcing. Uber systems stores and monitors
all the information about the users on each journey they take and such data are used to
determining demand, resource allocation, and set fares. After several trips, Uber will know
where you live, where you work, where you are traveling (Abrosimova K, 2014).
Use of Big data
Storing bulks of data does not help a business in any way. Business value for data is
created when the data is analyzed, made available to users, and the analysis used to make better
and informed decisions. Big data has improved the analytics as now there is, even more, data to
analyze.
Uber uses big data in the following ways:
Business strategy
Improve user experience to drive more profit to drive more profitable customer
engagements
Uber knows where returning customer lives, works or goes to hang out on the weekend and
the frequency in which the request Uber services. To improve the match result between a rider
and Uber drivers, Uber uses the ratings and feedback system. Uber customers are given a chance
to rate the drivers. Drivers also have an opportunity to evaluate their riders. This is done on a
scale of 1 to 5 with 1 being the lowest score and five the highest score. If a driver rates the
customer as little as 3, the Uber availability may be compromised for that rider (Forsythe, A,
Haenikel, D, Zha, M, 2017). On the other hand, a driver with meager ratings may be punished by

BUSINESS INTELLIGENCE USING BIG DATA
Uber. The score is used for future paring and to give the client the best experience. If a driver
rates a rider below 3, then they will never be paired again with that rider.
Business Initiative
Improve match results
Hyperscale is a platform that combines digital platform with location-based mapping
technology to connect rider with the driver in closest proximity and with lowest fares by use of
Google map.
To improve the match result between a rider and Uber drivers, Uber uses the ratings and
feedback system. Uber customers are given a chance to rate the drivers. Drivers also have an
opportunity to evaluate their riders. This is done on a scale of 1 to 5 with 1 being the lowest
score and five the highest score. If a driver rates the customer as little as 3, the Uber availability
may be compromised for that rider. On the other hand, a driver with meager ratings may be
punished by Uber. The score is used for future paring (Fisher E ,2016)
Increase customers
To make sure that they serve that demanding market, Uber has come up with diverse services
like UberPool, UberRush (a package courier service), UberChopper, and UberFresh.
Proper allocation of services using big data analytics
Uber helps to analyze public transport networks in-depth, and this allows drivers to know
which places have less access to buses and trains.
Enhance security using driver Selfies powered by Microsoft Cognitive Services
A system called "Constant ID Check" requires a few drivers to incidentally take selfies
before tolerating ride demands, to confirm character and keep drivers' records from being
bargained.
Outcomes and Critical Success Factors (CSF)

BUSINESS INTELLIGENCE USING BIG DATA
Optimize ratings and feedback system
Uber customers are given a chance to rate the drivers. Drivers also have an opportunity to
assess their riders. This is done on a scale of 1 to 5 with 1 being the lowest score and five the
highest score. If a driver rates the customer as little as 3, the Uber availability may be
compromised for that rider. On the other hand, a driver with very low ratings may be punished
by Uber. The score is used for future paring. If a driver rates a rider below 3, then they will never
be paired again with that rider (Forsythe et al, 2017).
Pricing value of the journey
In most cities all where Uber has its operations, payments and pricing are done in some
similar ways. The rider has the option to pay upfront where the fare is quoted to them. In some
cities, however, Uber doesn’t have upfront payments but instead calculates the price of the rider -
using a taximeter which chargers the rider using the time and the distance they travel.
On certain times Uber offers promotions to its customers on specific routes. At the end of the
ride, a rider makes payment with their preferred mode of payment which was pre-selected. This
could range from a credit card, mobile money methods are locally available or in some cities by
use of Google and Airtel wallets (Forsythe et al, 2017). After the ride is over, in some cities,
Uber offers the option where the rider can add a tip to the driver which is also billed using the
method the customer chooses.
Generate profile for passenger and driver using rating system
The use of rating and feedback system where passengers rate drivers and vice versa. If the
ratings are low, the action is taken to punish the driver, and in the case of a customer, then they
might lose the Uber services. These ratings are also used to pair a controller and a rider for future
trips. If the ratings are low, then such a customer and the driver never gets matched again.
Allocate resources by analyzing public transport networks.

BUSINESS INTELLIGENCE USING BIG DATA
Uber uses traffic data to leverage and adjust its prices. Traffic data is used to come up with
traffic patterns which then uber use to advice its drivers to strategically position itself. When the
traffic is high, then more drivers are directed to that spot to ease in responding to riders requests.
Use business intelligence to develop algorithms for monitoring traffic data and journey
time in real time.
Uber have algorithms that are able to calculate the price of a given trip depending on several
factors like; distance, time taken and traffic. These algorithms are designed in a way that they
will give you an estimate at the start of the trip and then when paying at the end of the trip, the
price might decrease or slightly increase.
Evaluate the match regarding proximity and fare
Uber drivers advised according on what routes to go and what routes have more passengers
making requests. One thing about Uber taxi is that the taxi is always moving and when a
customer makes a request, they get paired with the driver who is nearest to them.
Implement cashless system for payment integration
Uber has integrated some payment methods into its systems. Users can make payments by
use of Debit, credit cards and MasterCard. In some regions, a rider can pay using Paytm. In
countries like India, Uber has partnered with banks like NCPI to facilitate them with instant
money transfers. In other regions like Nairobi in Kenya, money operators Safaricom aids
payment of Uber services through its popular product M-Pesa.
Analyze and Implement market strategies like referrals, reviews, partnership, a loyalty
program (Uber VIP) and an omnichannel approach.
Right up 'til today, referrals are a necessary piece of Uber's advertising system. At the point
when an organization actualizes referral software, abruptly these viable yet hazy interchanges
end up noticeably unmistakable. They can boost promoters and first-time clients, scale their
showcasing, and pinpoint what's working and what isn't (Wiley, 2017).

BUSINESS INTELLIGENCE USING BIG DATA
Data sources
Uber, just like any other company collects big data from sources all around including Web
sites, Social media, Machine generated, RFID, Image, video, and audio and GPS
Driver data
Uber enters into a contract with drivers, and so a driver is taken as an independent contractor
and not an employee of Uber. Driver data such as gender, age, health, car age and type, ability to
drive and background information is collected and stored in Uber databases along with contact
details. It is not a must for Uber drivers to own either a car or a smart device, they can just rent
them out is they have met the minimum requirements to become Uber drivers.
Rider data
A rider is asked to sign up if they want to use the Uber services by providing their name,
email and phone number.
Payment data
Ones you have an account with Uber, you log in, and you can add your preferred mode of
payment. You can use a credit/debit card, visa card, MasterCard, local mobile money or any
other method that is available within your geographical region. Adding a new payment method,
e.g., a credit card will capture your name, CSV number, expiry date and the card number.
Traffic data
Uber uses traffic data to leverage and adjust its prices. Traffic data is used to come up with
traffic patterns which then uber use to advice its drivers to strategically position itself. When the
traffic is high, then more drivers are directed to that spot to ease in responding to riders requests.
GPS data
Uber manages billions of GPS locations. The Uber systems perform millions of transactions
in a single minute. Uber will use these GPS locations to manage better moving people and things
from place to place

BUSINESS INTELLIGENCE USING BIG DATA
Marketing data
Uber uses various data for carrying out its marketing objectives. GPS data, traffic data,
customer data and their requests are all analyzed to observe for to observe for trends.
Mobile data
Data used by Uber is very dynamic. A driver will report carrying a client at point X and
in a few minutes, the same driver will be in location y. These data is managed by use of real-time
technology.
Uber Technology Stack
HTTP Server Technologies
Uber utilizes the use of service-oriented architecture in service delivery where a single
service can communicate with another necessary service. Uber have implemented local HAProxy
cases for them to route JSON over HTTP requests to the other services as the frontend web
server – NGINX proxy to servers in the backend. This model ensures that troubleshooting is easy
and that the system can be easily modified and advanced in future ( Norouzi, A., Sertbas, A., &
Zaim, A. ,2014).
Client libraries
Uber have created client libraries to help integrate it with the uber application systems. They
have used Java, Android, iOS and Python programming languages to implement these client
libraries.
Programming language
At lower levels of Uber technology app, Uber engineers have mainly used Python, Node.js,
Go, and Java programming languages. Python was meant to be used by everyone else except the
marketplace where Node.js is used.
With time Uber started using Go and Java languages to attain high performance of the
system. Java is preferred as it takes advantage of it operating in an open source environment and
its ability to be easily integrated with other technologies such as Hadoop and other crucial
analytical tools, Holmes, A. (2014). . Go language, on the other hand, ensures efficiency, runtime

BUSINESS INTELLIGENCE USING BIG DATA
up speeds and simplicity of the system. Tornado language has been used together with Python,
but Go’s native support for concurrency is ideal for most new performance-critical services.
Server Libraries
Uber utilizes Bedrock web server which is built onto of Express.js framework. The
framework has a middleware which is default set for internationalization, to provide security and
ensure integration of other technologies into the system infrastructure, (Cohen, P et al. 2016).
Atreyu handles Uber’s internal server communication with the backend services while at the
same time integrating with Bedrock web server. Atreyu works similar to Relay as it allows Uber
to make requests to SOA service APIs with ease.
Databases
Previously, Uber used a consistent backend application written in Python that
used Postgres for data persistence. Over the years, this architecture has evolved to make a model
that is comprised of microservices and data platforms. Currently, Schemaless have replaced
Postgres. Schemaless is a sharing layer that is built on top of MySQL.
Data stores
Uber uses LevelDB for storing its data. The backend uses MySQL and standard Schemaless.
However, there are efforts to do away with MySQL and fully implement standard Schemaless.
Server Software
The Uber application requires the drivers to have a smart device –smartphone or tablet, and
clients must also have access to a smart device or use the web portal.
Operating System
Uber systems are cross-platform. This has been made so by making sure that you can access
the system via a smart appliance, i.e., a laptop, a PC, Android smart devices, iOS devices among
many other platforms.

BUSINESS INTELLIGENCE USING BIG DATA
For iOS, Uber stores the code in a monorepo that is built with Buck. When it comes to
component and sizing of iOS code, Masonry and Snapkit serve this purpose. KSCrash is used for
iOS crash detecting and reporting. Testing is both done by use of OCMock for Objective-C while
protocols are used to generate mocks in Swift.
For Android users, Gradle is used to build the system. OkHttp, Retrofit, and Gson for
networking functionalities while Dagger is utilized as the dependency injection framework. Open
source libraries have been implemented to keep the UI code concise and easy to interact with the
system. Picasso is used for image loading while Butter Knife bids views and callbacks to
methods and fields by use of annotation processing. Timber is used for logging and RxJava for
simplification of asynchronous and event-based programming, Marr, B. (2015).
Fig: Uber Technology Stack
Schemaless is a data store that Uber technologies have customized and used as its data store. It is
like a layer that is built on top of MySQL database.
Kafka is used for data streaming alongside production databases
Hadoop Distributed File System (HDFS) – It splits files into large blocks and distributes them
amongst the nodes in the group. During processing of data, HDFS sends the code to the nodes
that have the required data, and the nodes then process the data in parallel.
Hive, MapReduce, and HDFS – These are used together for data storage and to perform various
operations.

BUSINESS INTELLIGENCE USING BIG DATA
Tasks
Role of socio media in decision-making
Over the years, companies such as Uber have benefited enormously in using socio media
platforms to their advantage. They have tailored their socio sites in such a way that they are
based on a particular region (Pînzaru, F., Zbuchea, A., & Vidu, C. ,2016). Uber has their
presence in all significant and famous socio networking sites, i.e., Facebook, Instagram, twitter,
snapchat, LinkedIn just to mention a few. The following are ways in which Uber have utilized
socio platforms: Marketing, Customer service, and Public relations.
Uber has maintained a steady engagement with its clients through socio sites. They have
used the sites to post their new partnership deals, launch new products and even customer
promotion. Uber puts up advertisements according to the region a user is located. They
will use a hashtag, e.g., # Nairobi, Enjoy the free ride to Jacaranda hotel. Customers in
Nairobi will be able to see this promotion and redeem them accordingly.
Uber uses sites like Twitter to launch its new products and services as here they can
control even the talk about their products and services by creating trending topics.
The most important part of any company is the customer. Uber has been able to reach the
clients and resolve their needs right from their socio sites.
Data Visualizing and Analytics
Visualization-based data discovery solution that offers highly interactive and graphical user
interface is built on in-memory architecture and are geared towards addressing business users'
unmet ease of use and rapid deployment needs. These solutions enable the users to explore data
without much training, making them accessible to a more extensive scope of workers than
conventional business analysis tools (“Kapstone L/c, nd”).
Collecting data is one step in big data and business intelligence. Uber will now compose the
same data, analyze it and interpret it to use it for decision-making. For instance, Uber manages
billions of GPS locations. The Uber systems perform millions of transactions in a single minute.
Uber will use these GPS locations to manage better moving people and things from place to
place(“ data visualization,” n.d.).

BUSINESS INTELLIGENCE USING BIG DATA
In Uber, data visualization specialists range from information design to graphics
professionals. Uber manages all the information from mapping and system developments to data
that the public sees like drivers profiles. Most of Uber’s data visualizations have not been done
before, and thus these visualizations require that Uber come up with in-house development tools,
(Marr, B. et al, 2015).
MDM to support DS&BI
Master data is not connected to facts and is usually characterized as dimensions in
Business intelligence systems. Data analytics and Master Data management has enabled Uber
make a very enormous transformation in how they carry out their daily transactions. In Uber,
MDM input data are defined using the default data names and data definitions. Uber benefits
from Master data use in Business Intelligence as it is able to reuse both similar data names and
data definitions across all its micro services modules (Holmes, A, 2014). Consequently, the use
of similar data names and definitions guarantees consistency of all Uber data.
The use of Master Data management also helps in the reading, analyzing and
interpretation of Business Intelligence scorecards, reports, OLAP, dashboards among other BI
features.
Support of NoSQL for Big Data Analytics
NoSQL has been preferred over SQL-based relational databases due to its increased
scalability, more customization features, flexibility and its support for dynamic representation
designs. NoSQL enables enterprises to collect, analyze, store and interpret Big data even if it is
non-uniform data. NoSQL gives a company the merits of Big data search, capture, storage,
transfer, analysis and sharing of these same data.
NoSQL has removed the monopoly that was enjoyed by SQL-based databases. However,
it doesn’t mean that a company using NoSQL may not need the use of SQL database anymore.
Furthermore, NoSQL database supports the use of some SQL structures. NoSQL is a perfect fit
for cloud-based web applications. It is also a great fit when it comes to companies that are using
big data analytics and business intelligence tools such as Uber (Bulger, M., Taylor, G., &
Schroeder, R. , 2014).

BUSINESS INTELLIGENCE USING BIG DATA
The need for a technology that would support cloud computing, web, big users, big data
and data analytics necessitated the need to come up with NoSQL. This is an alternative database
that does not require the need to use fixed schemas. With this kind of database, big companies
can now scale horizontally without having to do major join operations on data.
How Uber has used NoSQL Databases in Big Data
NoSQL has been preferred over SQL-based relational databases due to its increased
scalability, more customization features and its support for dynamic schema designs. Uber for
one, use NoSQL as it enables them to collect, analyze, store and interpret Big data even if it’s not
uniform data.
Uber choice of NoSQL is based on the kind of big data they handle. NoSQL can handle
big data, support cloud computing and is an excellent fit for use alongside Business intelligence
systems as well as Data analytics. NoSQL gives Uber the merits of Big data search, capture,
storage, transfer, analysis and sharing of these same data (Turner, D., & Gorichanaz, T. ,2016).
The need for a technology that would support cloud computing, web, big users, big data
and data analytics necessitated the need to come up with NoSQL. This is an alternative database
that does not require the need to use fixed schemas. With this kind of database, Uber can now
scale horizontally without having to do major join operations on data.
Big Data Value creation process
One of the huge effects of big data is the hierarchical change or change necessary to help
and adventure the big data opportunity. Old positions should be reclassified, and new roles
presented, making both open doors and anxiety for people and associations alike (“Chapter 4:
group impact of big data, nd”). Uber have used big data and business intelligence to capture
roles, responsibilities, and expectations of its customers.
For instance, the use of rating and feedback system where riders rate drivers and vice
versa. If the ratings are low, the action is taken to punish the driver, and in the case of a
customer, then they might lose the Uber services. These ratings are also used to pair a driver and
a rider for future trips. If the ratings are low, then such a customer and the driver never gets
matched again, Marr, B. (2016).

BUSINESS INTELLIGENCE USING BIG DATA
Uber understands the linkage between its data and what the world needs. After they
collect all the data, they store it and analyze it to predict some things, e.g., Customer waiting
time, recommending where drivers should place themselves via heatmap to benefit from areas
with best prices and most clients (“Uber and big data,” n.d.). The best thing is that Uber
implements all these in real-time for both their drivers and the riders.
Conclusion
To conclude, Uber operates with mostly the same dynamics as other taxi operators but
makes its users feel like they have a personal driver in the safety of their vehicle. Uber markets
people with real purchasing power, among ordinary users have multiple cars and are in the
workforce. This makes the statement “people don’t want cars, they want rides applicable in the
Philippines. The survey concludes that Uber has better services and creates a negative perception
on the taxicab. Uber has an edge in safety through effective information dissemination,
convenience through technological advancements in booking and GPS, and comfort through
newer cars and performance conscious drivers. The only clear disadvantage of Uber is its surge
pricing feature. Aside from that, commuters have embraced Uber. Assuming both modes have
the same price and service, most users will still prefer Uber

BUSINESS INTELLIGENCE USING BIG DATA
References
Abrosimova, K 2014, ‘Building an App Like Uber: What Is the Uber App Made From?’, blog post,
22 May, viewed 09 September 2017, https://medium.com/yalantis-mobile/Uber-underlying-
technologies-and-how-it-actually-works-526f55b37c6f
Bulger, M., Taylor, G., & Schroeder, R. (2014). Data-driven business models: challenges and
opportunities of big data. Online], Oxford Internet Institute,
https://www.oii.ox.ac.u
k/archive/downloads/publications/nemode_business_models_for_bigdata_2014_oxford.pdf
Chapter 4: Organizational Impact Of Big Data - Big Data ... (n.d.). Retrieved from
https://www.safaribooksonline.com/library/view/big-data-understanding/9781118740
Cohen, P et al. 2016, ‘Using Big Data to Estimate Consumer Surplus: The Case of Uber.
Cohen, P., Hahn, R. W., Hall, J., Levitt, S. D., Metcalfe, R., & National Bureau of Economic
Research,. (2016). Using big data to estimate consumer surplus: The case of Uber.
Fisher, E 2016, ‘Uber’s Marketing Strategy in 7 Steps, Without the Bad Press’, blog post, 28 January,
viewed 11 September 2017,http://www.annexcloud.com/blog/2016/01/28/Ubers-marketing-
strategy-in-7-steps/
Forsythe, A, Haenikel, D, Zha, M 2017,Redesigning Uber Engineering's Mobile Content Delivery
Ecosystem, viewed 10 September 2017, https://eng.Uber.com/tech-stack-part-one/
Forsythe, A, Haenikel, D, Zha, M 2017, Redesigning Uber Engineering's Mobile Content Delivery
Ecosystem, viewed 10 September 2017, https://eng.Uber.com/tech-stack-part-two/
Holmes, A. (2014). Hadoop in practice.
Henke, N et al. 2016, The Age of Analytics: Competing in a Data-driven World, Mckinse Global
Institute
Kapstone Llc - - Big Data. (n.d.). Retrieved from http://kapstonellc.com/our-solutions/big-data/

BUSINESS INTELLIGENCE USING BIG DATA
How Uber Uses Data to Improve Their Service and Create the New Wave of Mobility. (n.d.).
Retrieved October 3, 2017, from https://blog.kissmetrics.com/how-Uber-uses-data/
Marr, B 2015, ‘The Amazing Ways Uber Is Using Big Data’, blog post, 7 May, viewed 11 September
2017, http://www.datasciencecentral.com/profiles/blogs/the-amazing-ways-Uber-is-using-big-
data
Marr, B 2016, Uber: How Big Data is at the Centre of Uber's Transportation Business, in Big Data in
Practice: How 45 Successful Companies Used Big Data Analytics to Deliver Extraordinary
Results, John Wiley & Sons, Ltd, Chichester, UK.doi: 10.1002/9781119278825.ch42
Marr, B. (2016). Big data in practice: How 45 successful companies used big data analytics to
deliver extraordinary results.
Microsoft 2016, How Uber is using driver selfies to enhance security, powered by Microsoft
Cognitive Services, video, 3o September, viewed 16 September 2017,
https://www.youtube.com/watch?v=aEBi4OpXU4Q
Norouzi, A., Sertbas, A., & Zaim, A. (2014). Evaluation Of Effective Parameters Of Energy
Consumption In Wireless Sensor Networks. International Journal of Computer Research, 21(1),
73.
Pînzaru, F., Zbuchea, A., & Vidu, C. (2016). Exploring Challenges for Managers in the Digital
Economy: Working Paper. European Conference on Management, Leadership & Governance,
328
Skill Pages - Apache Hadoop | Dice.com. (n.d.). Retrieved from
https://www.dice.com/skills/Hadoop.html
TechStacks 2016, Uber, viewed 10 September 2017, http:// techstacks.io/Uber
Turner, D., & Gorichanaz, T. (2016). Old Skills and New Practices Mean Radical Change for Library
Education. Journal of Education for Library and Information Science, 57(3), 239.

BUSINESS INTELLIGENCE USING BIG DATA
Uber | Social Media Marketing. (n.d.). Retrieved October 3, 2017, from
http://www.postcontrolmarketing.com/453/2016/07/31/Uber/
Wiley: Big Data In Practice: How 45 Successful Companies ... (n.d.). Retrieved from
http://www.wiley.com/WileyCDA/WileyTitle/productCd-1119231388.html