Moving Toward Big Data: Challenges, Trends and Perspectives

International Journal of Research in Engineering and Science (IJRES)
ISSN (Online): 2320-9364, ISSN (Print): 2320-9356
www.ijres.org Volume 5 Issue 1 ǁ January. 2017 ǁ PP.16-21
www.ijres.org 16 | Page
Moving Toward Big Data: Challenges, Trends and Perspectives
Ming-Hsing Chiu1
, Joshua Mcadams2
1
Department of Computer Science, Dillard University, New Orleans, LA, USA
2
Google-in-Residence Scholar, Dillard University, New Orleans, LA, USA
Abstract: Big data refers to the organizational data asset that exceeds the volume, velocity, and variety of data
typically stored using traditional structured database technologies. This type of data has become the important
resource from which organizations can get valuable insightand make business decision by applying predictive
analysis. This paper provides a comprehensive view of current status of big data development,starting from the
definition and the description of Hadoop and MapReduce – the framework that standardizes the use of cluster of
commodity machines to analyze big data. For the organizations that are ready to embrace big data technology,
significant adjustments on infrastructure andthe roles played byIT professionals and BI practitioners must be
anticipated which is discussed in the challenges of big data section. The landscape of big data development
change rapidly which is directly related to the trend of big data. Clearly, a major part of the trend is the result
ofthe attempt to deal with the challenges discussed earlier. Lastly the paper includes the most recent job
prospective related to big data. The description of several job titles that comprise the workforce in the area of
big data are also included.
Keywords: Big Data, MapReduce and Hadoop, Data Scientist,
I. INTRODUCTION
Big data technology has revolutionize the way organizations collect, manage and utilize the
explosive amount of data that has been generated in recent years. According to the report by International
Data Corporation (IDC), from the beginning of time through 2011, the world contained nearly 1 zettabyte
(1021
bytes) of data. Then the floodgate opened. 2011 saw the generation of twice (1.8 zettabytes) the
amount of all-through-history number in just one year.It was estimated by IDC that by 2020, just 4 years
from now, we’ll be generating 35 zettabytes per year. This amount of data is clearly far exceeding that
traditional structured database technologies can handle. Owing to the development of new data processing
technology such as Hadoop and MapReduce, organizations are now able to utilize predictive analytics on
the vast amount of data to obtain insightful information. Many organization have been using the data asset
to gain the valuable insight that are not considered possible before.
Famous examples include: 1.IBM were using electronic health record data to predict heart
disease [4], and using data from the World Health Organization to predictthe location of future malaria
outbreaks [5]. 2. Target uses a wealth of customer data to predict future purchasing habits. Specifically,
pregnancy kicks off a chain of purchases that are fairly distinctive. Target’s data collection is spookily
prescient, sending one teen customer nappy vouchers before her own father knew she was pregnant [3]. 3.
WeatherSignal App works by repurposing the sensors in Android devices to map atmospheric readings
[6]. Handsets such as the Samsun S4, contain a barometer, hygrometer (humidity), ambient thermometer
and lightmeter. Obviously, the prospect of millions of personal weather stations feeding into one machine
that will average out readings is exciting, and one that has the potential to improve forecasting. 4. One of
the most exciting examples is using big data to predict the ongoing 2016 presidential election. OpenText’s
U.S. Election Tracker reads, analyzes and visualizes key U.S. Election big data, every day, from hundreds
of news sources, helping the U.S. voters make more informed decisions. The analysis includes natural
language processing, semantic processing, sentiment analysis and opinion algorithms [7].
In the following section, 5Vs of big data isdescribed. The frameworkof Hadoop and MapReduce
that enables the big data analytics is introduced in section 3. Followed by the discussion of challenges of
big data. Section 5 lists the trend of big data development. Section 6 covers the job perspectives related to
big data. Finally, section 7 contains some concluding remarks.
II. 5Vs OF BIG DATA
Big data is high-volume, high-velocity and high-variety information assets that demand cost-
effective, innovative forms of information processing for enhanced insight and decision making. Volume
describes the amount of data generated by various sources. Big data is usually associated with this
characteristic. Velocity describes the frequency at which data is generated, captured and shared. Variety
describes the diversity of data, which include structured and unstructured data. Structured data refers to

Moving Toward Big Data: Challenges, Trends And Perspectives
the type of data found in spreadsheet or in the relational database, which comprise only about 5% of all
active data. All other data are considered unstructured data,including but not limiting to audio, video, log
files, social media streams, and sensor data from the “internet of things”.To illustrate how fast data can be
generated, Twitter produces over 90 million tweets per day. And Walmart handles more than 1 million
customer transactions every hour, which are imported into databases estimated to contain more than 2.5
petabytes (2560 terabytes) of data, the equivalent of 167 times the information contained in all the books
in the US Library of Congress.
Veracity is the 4th
V coined by IBM, which symbolizes the untrustworthy nature in some sources of
data.Due to the integration of disparate data systems, the correctness and accuracy of information is often
questionable. Value, the 5th
V, was introduced by Oracle as a crucial characteristic of Big Data. The ability to
understand and manage the vast amount of data, and then integrate them into the larger Business Intelligence
ecosystem can provide previously unknown insights associated with the organizations.
III. MAPREDUCE AND HADOOP
The underliningengine that allows the quick processing of big data is the framework of Hadoop and
MapReduce. The concept of MapReduce was first proposed by Dean and Ghemawat from Google in 2004
[1],which became the primary programming framework for processing big data. It first splits the input data into
a number of chunks that is manageable by a single processor, then goes through the mapping, shuffling, and
reducing stages to get the final result.Apache Hadoop was implemented in 2006 by a group of software
developers headed by Cutting [2]. Hadoop is an architecture that enables MapReduce to be run on clusters of
commodity machines in parallel that are dynamically scalable. In other word, the number of machines can be
added or reduced on demand, depending on the load of data. The Distributed file system among the machines in
the clusters is called Hadoop Distributed File System (HDFS).
Table 1 depicts the size of data that would require the use of Hadoop. In 2009, Yahoo demonstrated the
power of Hadoop that ran 17 cluster of 24,000 machine, sorting one Terabyte of data in 62 second. In 2011,
Yahoo announced its new humongous Hadoop cluster, reaching 42K Hadoop nodes and hundreds of petabytes
of storage.
Table 1: Size of Data and Corresponding Content
IV. THE CHALLENGES OF BIG DATA
Conventionally, the relationship between IT (information technology) practitioners and BI (business
intelligence) professionals in an organization is relatively well defined: IT processes organizational data and
generatesstructured data such as relational database or spreadsheet, which then used by BI to make decisions.As
big data technology matures to a point in which more organizations are prepared to pilot and adopt it as a core
component of the information management and integrated analytics, IT and BI alike will bump up against a
number of challenges that must be addressed before any big data program can be successfully implemented.
Below is the discussion of the five most common faced challenges[8, 9].
4.1 Talent Gap Challenge
There is a growing community of application developers who are relying on leveraging Hadoop
MapReduce framework. The promotion of these technologies creates a group of experts that are experienced in
tool implementation and its use as a programming model, rather than the data management aspects. This

suggests that many big data tool experts remain somewhat immature on the practical aspects of data modeling,
data architecture, and data integration.
There is no doubt that as more data practitioners become engaged, the talent gap will gradually narrow
down, but when developers are not skilled at addressing these fundamental data architecture and data
management issues, the ability to achieve and maintain a competitive edge through technology adoption will be
severely impaired. Section 6 includes a description of all job titles that comprise the talents needed for the
complete big data operation.
4.2 Infrastructure/Platform Challenge
Big Data is a much talked about technology across businesses today. The majority of organizations are
convinced of its usefulness, however the implementations primarily focus on applications rather than
infrastructure. Most organizations have their own legacy computer systems. Some may have storage system
based on NetApp, EMC or other major players. Moving toward Hadoop environment may prove to be daunting
and challenging.
The organizations opt to take advantage of Big Data technology must answer a number of questions.
How do you apply big data to business applications? What kind of storage system and hardware architecture do
you attempt to build? Are you going to build it all by yourself? Unpredictability of business world must also be
taken into consideration. The underlying platform chosen must be robust and scalable in a cost effective manner.
Cloud service can be a viable solution for organizations choose to avoid installing their own infrastructures.
4.3 Data Sources Integration Challenge
The volume, velocity, and variety of big data has direct impact on data integration and management.
Organizations must first deal with the variety of data which may be generated by IT, Internet of things, or users.
It takes a lot of skills to put data in the right platform so that you can use it properly. For example, if the data
comes from social media content, you need to know who the user is in a general sense, such as a customer using
a particular set of products, and figure out what it’s you are trying to “visualize” out of the data.
Another aspect is due to the velocity of the data which make data synchronization necessary across the
data sources under certain circumstances. There are two types of synchronizations. From a data currency
perspective, it implies that the data coming from one source is not out of date with data coming from another
source. From a semantics perspective, synchronization implies the similarity of data concepts, definitions,
metadata, and the like. One of the promise of big data is that it’s possible to reduce the costly upfront data
preparation and data engineering which typically constitute 80% of the time and cost of data management.
4.4 The Governance Challenge
Data governance is the process of ensuring the compliance of organizational data policy, which can be
a difficult task considering the variety and reliability of the unstructured data. The objective of data governance
is to institute the right levels of control to achieve one of three outcomes: identify data issues that might have
negative business impact; prioritize those issues in relation to their corresponding business value drivers; and
have data stewards take the proper actions when alerted to the existence of those issues [10].
It is naive to assume that when it comes to big data governance the approach to data quality is the same
as traditional approaches. When we examine the key characteristics of big data analytics, the analogy with the
conventional approaches to data quality and data governance starts to break down.Big data applications look at
many input streams originating within and outside the organization, some taken from a variety of social
networking streams, syndicated data streams, news feeds, preconfigured search filters, public or open-sourced
datasets, sensor networks, or other unstructured data streams. Such diverse datasets resist singular approaches to
governance.
4.5 Big Data Syndication Challenge
Most of the practical cases for big data involve data availability, ranging from augmenting existing data
storage to providing access to end-users employing business intelligence tools for the purpose of value
discovery. These BI tools not only must be able to connect to one or more big data platforms, they must provide
transparency to the data consumers to reduce or eliminate the need for custom coding. At the same time, as the
number of data consumers grows, it can be anticipated a need to support a rapidly expanding collection of many
simultaneous user accesses. That demand may spike at different times or the day or in reaction to different
aspects of business process cycles. Ensuring right-time data availability and reusability to the community of data
consumers becomes a critical success factor.
Big data syndication provides a supply chain to the different types of downstream applications in a way
that is seamless and transparent to the consuming applications while elastically supports demands. Finally, it’s
important to realize that funding is a challenge that always plays a dominant role at every stage of big data
integration because these days most budget is very tight. As a result, every penny must be spent wisely. Clearly,

in establishing infrastructure, what we are looking for is predictability in cost. One of the reason developers of
big data move to Hadoop is because it’s complying with industrial standard hardware, and scalable. In addition,
the long term cost of managing the storage facilities, analytic platforms, and all the related infrastructure must
be taken into consideration.
V. THE TREND OF BIG DATA
The year 2015 was an important one in the world of big data. What used to be hype became the norm
as more businesses realized that data, in all forms and sizes, is critical to making the best possible decisions. In
2016 and beyond, we’ll see continued growth of systems that support unstructured data as well as massive
volumes of data [11]. These systems will evolve and mature to operate well inside of enterprise IT systems and
standards, enabling both business users and data scientists to fully realize the value of big data. A number of
trends are discussion below. Note that most of the trends included in the discussion are stimulated by the need to
tackle the challenges covered in previous section.
5.1 The NoSQL Take 0ver
We noted the increasing adoption of NoSQL technologies, which are commonly associated with
unstructured data, as the conventional relational database management systems become inadequate. Going
forward, the shift to NoSQL databases becoming a leading piece of the enterprise IT landscape as the benefits of
schema-less database concepts become more pronounced. According to Gartner’s Magic Quadrant [12], the
developers of Operational Database Management Systems, which in the past were dominated by Oracle, IBM,
Microsoft and SAP, have shifted to more than a dozen NoSQL developers, including MongoDB, SQLFire,
Casandra, HBase, and Amazon Web Services (with DynamoDB).
5.2 Apache Spark Lights Up Big Data
Apache Spark has moved from being a component of the Hadoop ecosystem to the big data platform of
choice for a number of enterprises. Spark provides dramatically increased data processing speed compared to
Hadoop and is now the largest big data open source project, according to Spark originator and Databricks co-
founder, MateiZaharia [13]. We see more and more compelling enterprise use cases around Spark, such as at
Goldman Sachs where Spark has become the major platform of big data analytics.
5.3 Hadoop continues to evolve
In a recent survey of 2,200 Hadoop customers, only 3% of respondents anticipate they will be doing
less with Hadoop in the next 12 months. 76% of those who already use Hadoop plan on doing more within the
next 3 months and finally, almost half of the companies that haven’t deployed Hadoop say they will within the
next 12 months.
As further evidence to the growing trend of Hadoop becoming a core part of the enterprise IT
landscape, we’ll see investment grows in the components surrounding enterprise systems such as security.
Apache Sentry project provides a system for enforcing fine-grained, role based authorization to data and
metadata stored on a Hadoop cluster. These are the types of capabilities that customers expect from their
enterprise-grade RDBMS platforms and are now coming to the forefront of the emerging big data technologies,
thus eliminating one more barrier to enterprise adoption. We also see a growing demand from end users for the
same fast data exploration capabilities they’ve come to expect from traditional data warehouses.
5.4 The Number Of Options For Preparing Data Grows.
As the demand to reduce the cost of preparing raw data for analytics mounts, a number of
developments have been underway. Self-service data preparation tools are exploding in popularity. This is in
part due to the shift toward generated data discovery tools such as Tableau that reduce time to analyze data.
Business users also want to be able to reduce the time and complexity of preparing data for analysis, something
that is especially important in the world of big data when dealing with a variety of data types and formats.
We’ve seen a host of innovation in this space from companies focused on end user data preparation for Big Data
such as Alteryx, Trifacta, Paxata and Lavastorm.
5.5 Data Warehouse Growth Is Heating Up In The Cloud
Data warehouse have not been phasing out as were expected. But it’s no secret that growth in this
segment of the market has been slow. But we now see a major shift in the application of this technology to the
cloud where Amazon led the way with an on-demand cloud data warehouse in Redshift. Redshift was AWS’s
fastest growing service but it now has competition from Google with BigQuery. Offerings from long time data
warehouse power players such as Microsoft (with Azure SQL Data Warehouse) and Teradata along with new
start-ups such as Snowflake, winner of Strata + Hadoop World 2015 Startup Showcase, also are gaining

adoption in this arena. Surveys indicate 90% of companies who have adopted Hadoop will also keep their data
warehouses. With these new cloud offerings, those customers can dynamically scale up or down the amount of
storage and compute resources in the data warehouse relative to the larger amounts of information stored in their
Hadoop data lake.
5.6 IoT, Cloud and Big Data Come Together.
The technology is still in its early days, but the data from devices in the Internet of Things will become
one of the “killer apps” for the cloud and a driver of petabyte scale data explosion. For this reason, we see
leading cloud and data companies such as Google, Amazon Web Services and Microsoft bringing Internet of
Things services to life where the data can move seamlessly to their cloud based analytics engines.
VI. JOB PERSPECTIVESRELATED TO BIG DATA
According to a Big Data Survey by Knowledgent, there are clear indicators that Big Data is moving out
of the experimental stage. Most respondents of this survey firmly believe that many organizations would be
utilizing Big Data in the production environment and over 60% of them responded that Big Data initiatives are
very or extremely important to their organizations. In 2015, 1.9 million jobs related to Big Data were created in
the United States to work on the IT-side of big data projects,but each of these jobs is supported by 3 new
jobs outside of IT.In general, several professionals with different skills are required to complete any project. For
example, as mentioned earlier,around 80 % of effort for big data project are spent on cleaning, integrating, and
transforming the data before it can be utilized.
EMC, IBM, Cisco, Oracle are just a few of the top companies who are looking for Big Data skills.
Table 2 is a distribution of job demands of the top Big Data employers today, according to Forbes [14]
Table 2: Big Data Jobs Demand from Major Employers
Jobs that require big data expertise pay well, too. The average salary, according to a report by Forbes last year,
is $124,000. Below is the list of 7 most common job titles related to big data [15].
Data Scientist
Sometimes referred as “the sexiest job in the 21 century”, data scientists are responsible for analyzing
data and extracting useful information out of it.Data scientists thrive on solving real world problems with real
data. They are very good at using different techniques for analyzing data from different sources to help business
make intelligent decisions. They need to have both skills of a software engineer and an applied scientist.
Data Analyst
The data analyst collects, processes and performs statistical data analyses. A data analyst needs to
understand R, Python, HTML, JavaScript, C/C++, SQL and NoSQL databases.
Data Architect
Data architects are responsible for the design, structure, and maintenance of data, ensuring the accuracy
and accessibility of data.
Data Engineer
Big data engineers are all-purpose workforce of a big data analytics operation. They often come from
programming backgrounds, and are experts in big data frameworks, such as Hadoop. They’re called on to ensure
that data pipelines are scalable, repeatable, and secure, and can serve multiple constituents in the enterprise.

Statistician
Collect, analyze and interpret qualitative and quantitative data with statistical theories and methods.
Statisticians have to be mathematicians, who can churn out reports basis some data provided to them. The
second part is about being able to use a big data technology to automate data processing that provides insight on
a real-time basis.
Database Administrator
Big data and databases go hand-in-hand. A database administrator ensures that the database is
available, is performing properly, and is secure.
Business Analyst
Business analysts are primarily responsible for generating insights that convert big data into business
value. This can include translating insights and patterns embedded in data assets into language understood by
the business administrator.
VII. CONCLUSION
Big data technology has revolutionize the way organizations collect, manage and utilize the
explosive amount of data that has been generated in recent years, including but not limiting to audio,
video, log files, social media streams, and sensor data from the “internet of things”. This paper provides a
comprehensive view of current status of big data development. This paper has described and discussed the
5Vs of big data, the framework of Hadoop and MapReduce that enables the big data analytics, the
challenges of big data, the trend of big data development, and the job perspectives related to big data. This
paper will prove most beneficial to the planner of the organizations ready to adopt big data technology as
well as for the newly grads that are looking for jobs in the area of big data.
REFERENCES
[1]. MapReduce: Simplified Data Processing on Large Clusters, Dean &Ghemawat. OSDI 2004, San
Francisco, CA, December, 2004.
[2]. Cutting, Doug (28 Jan 2006). "new mailing lists request: hadoop". issues.apache.org.
[3]. http://www.nytimes.com/2012/02/19/magazine/shopping-
habits.html?pagewanted=1&_r=3&hp&pagewanted=all&
[4]. http://cloudtimes.org/2013/10/14/ibm-wants-to-predict-heart-disease-through-big-data-analytics/
[5]. http://venturebeat.com/2013/09/29/ibm-uses-big-data-to-predict-outbreaks-of-dengue-fever-and-
malaria/#3QW6mxLOYOPrZTpC.99
[6]. http://blogs.scientificamerican.com/guest-blog/weathersignal-big-data-meets-forecasting/
[7]. http://www.forbes.com/sites/adigaskell/2016/03/11/using-big-data-to-predict-the
election/#c07b7b776642
[8]. https://www.sas.com/resources/asset/five-big-data-challenges-article.pdf
[9]. https://www.progress.com/docs/default-source/default-document-
library/Progress/Documents/Papers/Addressing-Five-Emerging-Challenges-of-Big-Data.pdf
[10]. http://data-informed.com/data-governance-for-big-dta-analytics-considerations-for-data-policies-and-
processes/
[11]. http://www.tableau.com/about/blog/2015/12/top-8-trends-big-data-2016-47846
[12]. https://www.gartner.com/doc/reprints?id=1-2XXUR6C&ct=160204&st=sb
[13]. http://spark.apache.org/
[14]. http://www.forbes.com/sites/louiscolumbus/2015/11/16/where-big-data-jobs-will-be-in-
2016/#438e0339f7f1
[15]. http://talkincloud.com/cloud-computing/7-big-data-jobs-you-need-know#slide-7-field_images-54391

Moving Toward Big Data: Challenges, Trends and Perspectives

More Related Content

What's hot (20)

Similar to Moving Toward Big Data: Challenges, Trends and Perspectives (20)

More from IJRESJOURNAL (20)

Recently uploaded (20)

Moving Toward Big Data: Challenges, Trends and Perspectives