Sentiment Analysis: Big Data, Data Collection, and Storage

This report provides an executive summary and detailed analysis of big data sentiment analysis, focusing on data collection methods, storage systems (including No-SQL databases), and consumer-centric product design. The report examines how businesses, such as Woolworths, utilize big data to understand customer needs, improve product experiences, and gain a competitive edge. It discusses the importance of data collection through social media and other platforms, storage solutions like relational and No-SQL databases, and consumer-centric design principles to meet customer demands. The report also highlights the significance of recommendation systems and the impact of big data on decision-making in business, emphasizing the shift towards technologies like Hadoop and cloud solutions. The report concludes by outlining how businesses can use these insights to enhance customer satisfaction and improve sales.

Big Data for Sentiment Analysis
Executive summary
Big data is defined to be a large volume of data that are complicated and cannot be
analyzed or efficiently handled and processed using traditional tools. There are different
types of big data out which there is structured, semi-structured and unstructured data.
Depending on each type of data collected, they will be handled using different types of
data bases. For instance, structured data are handled using relational databases while
unstructured data are handled using non-relational databases. Big data are collected
using google, social media e.t.c. and then directed to the databases for storage. Data
from outside sources are first extracted and changed by ELT before they are loaded into
the databases. For proper management of unstructured data that are collected from
various sources including social media are stored in No-SQL databases. Such
databases are always focused to carry out massive scaling, simplify application
development and data model flexibility. Consumer centric design is the process of
setting goods and services towards the needs, wants and demands on the side of the
consumers both as a result of designing and improving the quality of product, service
and even the contents in order to fulfil customers’ product experience. Recommendation
system is important for online business and for the continuity of business in case of
power outage is by using power backup.
Introduction
Big data is the enormous volume of data that are complicated and cannot be analyzed
or efficiently handled and processed by traditional tools. Depending on how data are

collected and the tools that were used to collect data, they keep on streaming in and
piling up from time to time for storage. Since reaching the customers in their various
places physically is difficult, companies and business organizations use social media
and other platforms to collect data in order to meet the demands of their customers.
Due to this therefore, big data sentiment analysis prove to play important role in
organizing and exhaustively deriving the meaning of the collected data for decision
making in business (Cambria et al, 2013). Unlike small and medium data, big data cover
a wider and detailed range of customer requirements that can be important in solving
most of the problems facing the customers outside there about the products produced
by a particular business. Interest of adopting the use of big data by various business
organizations have grown in the past few years since its emergence (Chen et al, 2012).
As a result of this, many technologies such as the Hadoop, Real-Time solutions, Cloud
Solutions and many more have come to light and are gaining popularity with the
continued rise of big data adoption by business organizations (Kunzea, 2015). Change
in data collection tool has also resulted to change in the type of big data collected. Tools
like social media on the internet collect large volume of unstructured data that are then
organized by No-SQL databases for storage (Wu et al, 2014). Woolworths is one of the
companies in Australia that put big data into practice where they spend heavily on it to
enjoy its fruits. The supermarket is currently facing a slow market growth as a result of
competition in the market. Venturing into big data was now aimed at capturing all the
information from customers through the social media that could help them handle the
problems that are currently there or might be detected by the data in the future.

Discussion of topics
Data collection system
With reliance on Quantium as one of the Australia’s oldest and biggest data company,
Woolworths assigned them to collect big data and help in carrying out big data
analytics. The Quantium Company collects data from the field through the internet
portals such as google and social media about Woolworths’ products from their
customers. Data that are mostly collected through such data collection tools are
unstructured where the users of the social media just come up with any issue
concerning the business that was not structured or even taken care of in the earlier
preparation of the data (Tufekci, 2014). The said portals where the customers drop their
comments about the products that are sold by the business (Woolworths) will be
streamed to the business’ database for storage and can be retrieved at any time in
future for analysis and interpretation. Once the data have been collected, the specialists
handling big data will therefore continue further to make use of the data through
carrying out big data and sentiment analysis.
Storage system
Where and how big data are stored stands to be a big question to all those who wish to
adopt its use. Retrieval of structured data through traditional methods include rational
databases, data warehouses etc. collected data are taken from operational data stores
where they are uploaded using Extract Transform Load (ELT) which are responsible for

extraction of data from outside sources and change the data to fit the needs in operation
thereafter loading data into the databases (Katal et al, 2013). Before the data is made
available for data mining, the data is cleaned, transformed and made ready for
analytical functions (Sagiroglu, 2013). The traditional Enterprise Data Warehouse
(EDW) do not allow the incorporation of new sources of data until when the data is
cleansed and integrated. Irrespective of quality of data, all data sources are attracted by
big data environments since they are made magnetic. Storage of big data should be in a
flexible way in such a way that data can rapidly be adapted and easily produced by the
analysts (Herodotou et al, 2011). Due to this, agile database is needed where its
physical and logical contents can be easily adapted with fast data evolution. The study
and analysis of enormous dataset by the analysts require the incorporation and use of
complicated statistical methods to excavate data up and down. Deep big data repository
also need to be done in order to have sophisticated algorithm runtime engine served.
For proper management of unstructured data that are collected from various sources
including social media are stored in No-SQL databases (Moniruzzaman and Hossain,
2013). Such databases are always focused to carry out massive scaling, simplify
application development and data model flexibility. Data management and storage are
separated in No-SQL, this is contrary to relational databases that manage structured
data. Tasks of managing data that are to be written in specific database languages are
written in the application layer since such databases focus much on high performance
scalable data storage (Moniruzzaman and Hossain, 2013). Elimination of disk
input/output enables real time response from the database since data in server memory
is managed by in-memory databases. Silicon based main memory are used for storage

of big data instead of using mechanical disk drives hence boosting performance
magnitude allowing development of new applications (Plattner and Zeier, 2011).
Consumer Centric product design
This is the process of setting goods and services towards the needs, wants and
demerits on the side of the consumers both as a result of designing and improving the
quality of product, service and even the contents (Tseng and Piller, 2011). Such
process is seen to put designers into task since they have to critically think and analyze
how the consumers (customers) will be able to interact with the products and also try to
verify the assumptions of the process at each and every stage. In the process, when the
designers are looking and testing into each and every stage, some few adjustments will
be required on both the products and the management techniques (Tseng and Piller,
2011). All these are done with the aim of optimizing and meeting the goals of the
products through building how the customers will be interacting with the products
instead of coming up with the products that will need the customers to adjust their
behaviors towards fitting to the requirement of the products. Huge returns can be
expected from consumer centric design since it is much onto the customers’ satisfaction
therefore goods and services are made available to meet all the customers’ requirement
which will boost the sales thus maximizing the profit of the business since many
customers will be attracted towards the goods. This design process matters a lot since
the customers have control over what they want than ever before since they can be
having the related and alternative products that come through widespread of internet
that is to the reach of almost everyone in the world today (Pang et al, 2015). Customers

can therefore perform google search to obtain goods they want incase what they have
at their reach do not meet their demands and this is what the consumer centric design is
trying to work hard to minimize at all cost since the businesses want to keep their
customers.
Above all, consumer centric design is not only concerned with the good services you
provide to the customers through goods and services you offer them but also the
experience of the customers in the post purchasing process of that same good and
services (Tseng and Piller, 2011). Through the collected data and carrying out
sentiment analysis, the analysts are supposed to use analytical techniques to predict
customers’ needs. This will enable the business production team come up with goods
and services that bombast the customers’ experience with what they did not expect to
have in the market but will suddenly have close touch and experience with the product.
In that case therefore, consumer centric product design can be achieved through the
following practices. Firstly, the belief that customer comes first before anything else and
thus ensuring that good brands that are concerned with fulfilling the customer needs.
This perception makes the business to value the customer and see the world through
their customers and attaching themselves to a belief that there is no success without the
customers (Zomerdijk and Voss, 2010). Customer data are used to capture the
demands of the customers as revealed from big data sentiment analysis by the big data
analysts in order to achieve consumer centricity. Enough research and analysis of the
big data obtained from the customers towards their experience to the products and their
expectations of the products, brands of goods and services that are produced are
towards meeting the wants needs and also ensuring that all other goods and services

are produced in consideration to those wants and needs. Once the needs and wants of
the customers are met by the produced products, this will tap the consumers to stick to
the company’s products. Due to this therefore, in order to reduce the current problems
and deteriorating market, Woolworths is supposed to build its products towards the
customers’ needs and wants for it to stay competitive in the market.
Furthermore, in order to maximize customers’ product and service experience, the
business needs to produce brands that are concerned at building the relationships with
the customers (Tseng and Piller, 2011). Once the customers have good relationship
with the products of a particular business enterprise, it will be much easier for them to
become users of the products and services thus improving the sales. This is only
possible if the stored collected data are processed and analyzed for the best of the
business. As well, the brands that are aimed at coming up with customer centricity, they
need to analyze, come up with proper plans and only implement the customer
formulated strategies in order for the price to stay loyal to the customer and for the
profitability of the products in the business (Zomerdijk and Voss, 2010). Being that profit
is always the aim of each and every business, implementing strategies that will scare
away the customers will pull down the number of customers that interact and buy the
products hence reducing the sales thus having same effect on acquired profits in
business. The brands are therefore to ensure for loyalty of the customers.
Recommendation system
This is the extensive class that is obtained in the Web application that encompasses the
speculation of web users’ reactions to choices. One of the common examples of
recommendation system is; by basing on the interest of online readers, you decide to

offer them news articles. This focusing can only be with adequate research results
considering their past history of online reading (Davidson et al, 2010).
Various technologies are involved in recommendation system that can be categorized
into two major categories i.e. content based systems and collaborative filtering systems.
Content based system work by evaluating the feature of the recommended items while
collaborative filtering systems work by suggestion an item recommendation majorly
basing on commonality measures between the items and users. In such a case, when a
user is in demand of a certain product, he/ she will have the recommendation similar to
that of other users with similar request or demand (Sharma and Gera, 2013).
Recommendation system has various applications that can be applied in business. One
of them is product recommendations. Recommendation systems are widely used at
online retailers. Online venders are seen to struggle to exhibit their products to each
returning user of the system trying to suggest to them the products that they might be
interested in buying. The pick of suggestions to particular customers are not done at
random, but in regards to some purchasing decisions that were brought up by
analogous customers (Davidson et al, 2010). Another application recommendation of
movies. A good example for this is Netflix which do offer their customers with
suggestion of movies they might be interested in watching. The recommended movies
are always from the customer ratings, where the once with the highest ratings are
provided and recommended to the users to watch. Correct prediction of ratings is vital in
increasing the number of users who would take to the recommendations. Basing on the
past read articles, the news service providers try to dedicate some more articles that
might interest the readers. Before the recommendation of a particular news to read,

filtering is done to the news to ensure that there are similarity with the previously read
ones or there is match of words. This same principle is applied for the blog readers and
those who watch videos on the YouTube, they might do a search of a movie title by a
certain actor or actress then other options by the same actor or actress are displayed so
that one can trend with the movies one after the other.
Business continuity
In case of any disaster or power outage, the business should come up with various
plans to curb the situation and ensure that business operation do not come to a
standstill. Power outage is one of the serious problems that can limit the operation and
continuation of the online business. As a result, the business is supposed to point out
the systems that could have critical effect on the operation and continuation of the
business in advance (Omar et al,2011). On the same, symptoms of power problems are
supposed to be identified that could result to high risks of power outage. The checkup
should include voltage and current fluctuations and complains about lost data by the
users. Machines can be protected against brownout, blackouts and voltage dips using
UPS (Uninterruptible power supply). Standby power supply is an offline UPS that is
used to deliver power from the grid directly until when the power fails. Upon the failure
of power, battery powered inverter turns on and continues to supply power hence
ensuring the continuity operation of online business. Ferro-resonant transformer is used
to condition grid power by line interactive UPS. This type of UPS maintains steady
output voltage from varying input voltage and eliminates noise that could be produced
by motors and other devices that might be producing noise in the circuit. Though its
protection lasts for a short time, the protection is significant and adequate against

voltage variations. Above all, there is true UPS that is used to provide power under all
conditions and also ensures that its battery is charged thus ensuring the highest
protection to computing equipment. All the discussed different types of UPS offer power
protection to the equipment used in the operation of online business. Additionally, there
are private generators that should act as the power backup in case of the power outage
by coming on a short while after power outage. This option is always the best in case of
longer power outages that may go for hour or even days. Emergency power supply from
various sources are important in that they ensure for the operation of the business even
at the time of power outage since online business greatly depend on power for it
operation.
Conclusion and recommendations
We can therefore conclude that big data is important and useful to the operation of
business and by just collecting big data is not enough but should be followed with
expertise analytics that will help draw meaning from bid data. Depending on the type of
tool used to collect big data, different types of data are collected ranging from
unstructured, semi-structured and structured. Social media and google are found to be
major sources of big data thus making the collected data mainly being unstructured that
are best handled by special types of databases such as No-SQL. This database
ensures that data are stored and when it comes to data action, the data can easily be
accessible at a higher rate since they are stored in silicon based memory. Consumer
centric design is another important activity in business that will deal with improving the

products towards fitting the desires of the consumers so that they can have easy time
with the products in the market.
It can therefore be recommended from the above discussion that each business using
or aspiring to adopt the use of big data to higher the services of big data analytics
specialists. Also, being that most of the data collected are unstructured and from the
internet, it is important for the big data analysts to apply sentiment analysis in order to
draw meaning out of the data in their analysis. Since ordinary database tools cannot be
used to efficiently handle big data, databases such as No-SQL are supposed to be used
to easily handle big and unstructured data. Before the data analysis specialists start the
analysis process, they are supposed to first understand the data and have vast
knowledge with what the data represent. Recommended system is recommended for
the online business handlers where they can be suggesting to their customers with
similar desire a certain range of goods and services. For the continuous operation of
online business, the business operators are supposed to ensure that they have power
back up plans (i.e. generators, UPS, e.t.c.) to deal with power outage.

Reference
1. Chen, H., Chiang, R.H. and Storey, V.C., 2012. Business intelligence and
analytics: From big data to big impact. MIS quarterly, 36(4).
2. Davidson, J., Liebald, B., Liu, J., Nandy, P., Van Vleet, T., Gargi, U., Gupta, S.,
He, Y., Lambert, M., Livingston, B. and Sampath, D., 2010, September. The
YouTube video recommendation system. In Proceedings of the fourth ACM
conference on Recommender systems (pp. 293-296). ACM.
3. Katal, A., Wazid, M. and Goudar, R.H., 2013, August. Big data: issues,
challenges, tools and good practices. In Contemporary Computing (IC3), 2013
Sixth International Conference on (pp. 404-409). IEEE.
4. Moniruzzaman, A.B.M. and Hossain, S.A., 2013. Nosql database: New era of
databases for big data analytics-classification, characteristics and
comparison. arXiv preprint arXiv:1307.0191.
5. Omar, A., Alijani, D. and Mason, R., 2011. Information technology disaster
recovery plan: Case study. Academy of Strategic Management Journal, 10(2),
p.127.
6. Plattner, H., Zeier, A.: 2011. In-Memory Data Management: An Inflection Point
for Enterprise Applications. Springer, Heidelberg.
7. Pang, Z., Chen, Q., Han, W. and Zheng, L., 2015. Value-centric design of the
internet-of-things solution for food supply chain: Value creation, sensor portfolio
and information fusion. Information Systems Frontiers, 17(2), pp.289-319.