Chapter 2

Literature Review

Literature review creates a skeleton for research work. It is a base or foundation of research on
which whole research is dependent. In this chapter, we have done the literature survey in the
various areas of I-Commerce.

LITERATURE REVIEW 1

2.1 I-Commerce 2

2.2 Recommendation system in I-Commerce 6

2.3 Optimization Techniques in I-Commerce 9

2.4 Big Data in I-Commerce 22

2.5 Parallel Clustering Algorithms in I-Commerce 24

2.6 Prediction based Hybrid Optimization technique for I-Commerce 27

2.7 Comparative Analysis 35

A literature survey gives a brief idea about the research to understand innovation and can be trusted
by the audience. It created a blueprint for our research. Review of work plays an important role in
research as it helps to identify the gap in the research, with the various authors who have done the
work in the same field. This helps to enhance the knowledge and the methodologies used. It
contains existing knowledge of substantive findings with a theoretical and methodological
approach. The importance of the review is to know the work and methodology used by the
researchers.

I-Commerce is very popular all over the world and at a very nascent stage. Many studies are
conducted for I-Commerce. I-Commerce is in boom situation and witnessing a huge consumption
of products. At the same time, technological enhancement has been carried out in terms of the
capacity of online retailers, consumers. To increase the sell or demand of products many
researchers have developed new techniques like recommendation systems, optimization systems.
Several researched are done in this field; a literature review is divided into six parts.

2.1 I-Commerce
I-Commerce is an amalgamation of E-Commerce, M-Commerce and IoT Commerce. Many
researches have been carried out in the arena of E-Commerce, M-Commerce and IoT Commerce.
According to the Internet World Stats (2019), 4,38,38,10,342 internet users are active, out of which
Asia scores the highest number of users (50.1%) followed by Europe (16.4%), Africa (11.2%), and
remaining by other continents. Though the internet has been used by many users in the world,
internet access is not equal across the world.

Agarwal (2016) identified the impact of E-commerce on India. In this paper, authors have focused
on the categories of E-Commerce that is E-merchandise and E-Finance. According to author by
2020 India is expected to $100 billion online retail revenue out of which 35% is expected to
complete in upcoming year. In this paper author also focused on reviews of various research studies
on influence of E-commerce in India. E-Commerce is most trending field and most off the sell is
carried out by E-Commerce. So it is most important to conduct research for better sell of the
products.
Argade (2015) carried out to enhance the precision in the prediction of future mobile user behavior
in mobile commerce environment using MCP 1 (Mobile Commerce Predictor) framework. In order
to increase the prediction accuracy, this research has employed a new strategy called ISC
(Integrated Support and Confidence). The findings of this research revealed that the proposed
strategy has been provided with best accuracy and better performance when compared to existing
strategies like support only and integrated support and matching length. The concept of ISC2 was
to integrate the pattern confidence as well as support into the m-commerce behavior prediction.
The confidence has replicated the certainty of detected mobile commerce patterns.

Hammed (2014) directed an enhanced methodology for M-Commerce. This work presents an agile
traditional way for software engineering-based development methodology for M-Commerce. A
mobile application is used to developing the model for consumer acceptance and successful mobile
marketing. Mobile phones are a medium for personalized and direct customer communication.
Result indicates the guarantees future for 3G-Services in India. The main driver of successful
marketing is obtained risk and cost of adoption for M-Commerce.

Gitau (2014) identified different parameters that affect the M-Commerce parameters. In this
research, author has focused on trust between consumers and vendors. Building trust between
consumers and vendors is a major concern for mobile service providers. It leads to the adoption of
M-Commerce. Technical limitations and social behavior is the major reason for this adoption.
Many M-Commerce services are exists as of now like Amazon, Flipkart. At the same time it is
very important to define a system where it will relies on trust between consumers and vendors.

Hosting fact team (2019) depicts that there are 4.1 billion internet users are available. Out of which
Asia scores 49% and China has maximum number of internet users (802 Million). Maximum users
from China are mobile users. Iceland is a country in which 100% of users are using the internet.
In 2018 the internet influenced retail sale is 2.84 trillion dollars. In 2019 1.2 billion peoples are
expected to purchase online.

Jahanshahi (2015) emphasize on M-commerce activities carried out with handheld devices. The
work proves to necessary aspects and provides instructions of mobile application like travel, ERP,

1
Mobile Commerce Predictor
2
Integrated Support and Confidence
education, healthcare, Inventory, tracking, education and commerce. Many M-Commerce
applications are exists like Flipkart, Amazon, snapdeal, jabango etc. Mobile commerce survey is
explained. In India 811,589,101 out of 1210193422 are using a mobile phone that is 67.98 % of
population. Here, all E-Commerce operations are performed on mobile devices.

IT plays a vital role in the development of business sector and the financial sector. Khan (2016)
has studied and identified various challenges and benefits of E-commerce in emerging economy
due to increased use of handheld devices and internet. A new channel has been introduced for a
business transaction and online trading. He has also focused on benefits of E-commerce like it
deals with customer satisfaction, conveniences, ease of access, reliability. This study also predicts
the challenges of E-commerce.

Ly-Fie-Sugianto (2008) argues that revenue stream models are the main reason for adoption rate.
Four main factors influence M-Commerce adoption rate. Network-centric models, Basic M-
Commerce model, service content aggregator model and content provider service model are
focused on this research. The work has that affects the performed comparison between E-
Commerce and M-Commerce context. It also focused on different challenges in M-Commerce.

E-Commerce in India started in early 20’s, but due to lack of technological advancement e-
commerce is vanished in 2005. Again in 2019 E-commerce is started in a boom. Many
entrepreneurs have started their business online. Large changes are identified because of digital
economy. Lavuri (2016) studies the behavior of consumer and identified important parameters like
TV advertising, brand Image, secured transactions. All these parameters tend to customer
satisfaction.

Panigrahi (2016) analyze the impact of E-commerce on the textile industry. E-Commerce has
gained large popularity in last few decades. Diversify E-Commerce is implemented for textile and
apparel supply chain industries. However, risk is hanging between present customers and new
customers. Due to the ‘Digital India’ campaign started by our Hon’ble Prime Minister Narendra
Modi, value of E-commerce has increased. Also, remarkable transformation has taken place in
youth, many youths have started their own business and it leads to entrepreneurship. He has also
examines E-commerce ventures with various commercial and business models. It shows the drastic
changes in the field of E-commerce.
According to Prabhu (2018), the most important application of internet is E-Commerce. The main
motto of E-commerce is customer satisfaction. Since the lifestyle becomes easier and adaptive
because of online shopping, it is important for the business to hold and update the technologies as
per the ease of convenience for the consumer. IoT enabled devices can use internet for exchange
of data and helps in retailing and for the growth of Business. In this paper author also explained
different way that can be used to integrate IoT with E-Commerce is discussed.

I-Commerce plays a vital role in human life. Talha (2018) has investigated in their studies online
commerce will Touches Mountains in 2021. He has also emphasized on the role of E-commerce
and different parameters involved in Business success. Special strategies need to be implemented
in the future to increase the sales of products online. Continuous use of applications, E-Commerce
offers product and market opportunities. Due to the effect of globalization E-commerce has it owns
importance in the corporate world.

Xiaoming (2014) proposes IoT model for E-Commerce. Internet of thing is the latest technology
works with RFID, wireless network, sensing devices for communication. In this paper, author has
described technical limitations of thing developments, positioning, monitoring, management,
system configuration, etc. while implementing IoT in E-commerce, many problems are identified.
The author has also focused on real-time monitoring of the products, automation in households
with the help of IoT.

Zhang (2017) explores the usage of IoT in E-Commerce industry. In this paper, author has focused
on network infrastructure and E-Commerce perspective. How networking topologies are combined
with IoT to solve current problems in E-Commerce. I-Commerce is the most trending field in the
world. According to many statistics till 2021 whole world will transform in the era of internet
technology. As demand will be increased in the future, we need to consider various parameters
like buying, selling, recommendations, security, availability, efficiency and relevancy and most
important is customer satisfaction. Studying the above aspects and literature review, it motivates
us to study and perform research in the field of I-Commerce.

Above literature study for I-Commerce stated that, it is most important to conduct researches in
the field of I-Commerce. Till 2022 maximum products are available online and purchasing of the
products should be takes place through online portals. At the same time we need to think of various
parameters like buying, selling, recommendations, security, availability, efficiency and relevancy
and most important is customer satisfaction. To achieve this objective we need to develop a
recommendation system where user can buy the products.

2.2 Recommendation system in I-Commerce

Recommendation system in I-Commerce plays a vital role. It is the most important tool for I-
Commerce. In I-Commerce customer satisfaction is a most important parameter. To achieve this
objective many researchers have implemented different recommender systems. This helps
customer to buy a good product with satisfaction. Recommendation system recommends a product
by considering similar people’s test and preferences as per past behavior.

Cha (2019) figure out the effect of artificial intelligence affects customer preferences. In this study
author has considered parameters `category preference' and `attribute preference'. Products can be
differentiated by content and their properties like tangibility. By conducting four hypotheses with
above properties author reach to conclusion artificial intelligence can be used effectively for
recommending products in E-Commerce.

Greenquist (2019) explores the use cases to provide book recommendations by applying a
recommender system and algorithm. In first phase Greenquist-Kilitcioglu-Bari (GKB) framework
is used for end to end process for building fully functional RS. It is hosted on internet with RMSE
of 0.842. This process highlights data collection, preprocessing, evaluation, selection, creating
hybrid models. This approach creates an efficient book recommendation and can be used in various
areas like finance, marketing data matching application, E-Commerce.

Glue labs has studied various recommender systems and explained the system with various
approaches used by different E-Commerce websites. As a case study glue labs have considered
Amazon. Amazon is trying to recommend different products by increasing the average order of
current purchase. It works like an intelligent salesman, who tries to sell maximum products along
with one product and help the parental organization to increase their productivity. It also explained
use of Machine learning in recommender system. In this article, they have explained the content-
based recommendation system, collaborative filtering and hybrid filtering.

Goswami (2018) has focused on main objectives of e-commerce to expand the business, reduce
costs, facilitate access to information and promote customer service in terms of time and quality.
Large companies and organizations, both in the public and private sectors, initially responded by
adding economically marginal sites to their networks, which expanded their reach. Today,
electronic commerce and in particular mobile commerce is not a separate line of business, but part
of a wider and more continuous evolution. A setting can be defined as user interface attributes that
are computer controlled, often using an algorithm or user model based on user choice, habits,
demographic information, etc. in order to support personalization mechanisms.

Kwon (2014) stated that in the recommender system is such that the predictions are done as per
the situational existing service. The user will be able to rate the chosen apartment with respect to
the provided criteria’s on the web site. There is no need to register, the ratings and viewings are
provided through this code & are unique. The mobile commerce has increasing acceptations &
increasing dependencies based on the user requirements & advancement proposed in the
technologies. The mobile development area has increased gradually. The improvement recorded
is 61.3% according to the recent statistics.

Pandey (2019) introduced the task of product recommendation. The aggregated marketplace is a
platform where multiple E-Commerce websites can be viewed, studied and analyzed. Trivago is
the best example of the aggregated marketplace. The main objective of this research is to provide
users with different alternatives for E-Commerce websites. To integrate this vector space
retrofitting method is used in language processing. Recent recommendations are just due to the
clicked data and not on the basis of user reviews and feedbacks. In this paper, author has tested a
model on Samsung mall app.

According to Resnick (1997) RS is defined as “system where people provide recommendations as

inputs which the system then aggregates and directs to appropriate recipients.” Traditional
recommender systems are mainly used to recommend services, peoples and products to consumers.
A recommender system is a collaborative learning, as the second term refers to the particular
algorithm that is recommended. The development of digital computing and communication
technologies is a key issue in the field of e-commerce in continuous development.

Sethi (2017) presents a recommender system for a movie by using collaborative filtering to
recommend the most relevant product. A recommender system is an integral and most important
part of websites like Flipkart, Amazon, Netflix, etc. Recommender system does not help to
increase the productivity of a particular product, but it helps in bringing the relevant products into
the notice of customers. For example, customers who are purchasing mobile phone should also
purchase headphone and mobile cover. This product is recommended by recommender system to
the customer. In this literature author has implemented collaborative algorithm and then to
compare on the basis of time analysis also applied ant colony optimization and artificial bee colony
optimization algorithms.

Wang (2019) implemented a collaborative filtering algorithm to build a recommender system with
capacity of sharing experiences. But recommender system can be easily trapped by Mathew effect
which tends to change the recommending priority of products. In this case, users are already aware
of the products so performs would be degrades in finding new items. For solving this issue author
has done a survey in China market by applying novel recommendation system. Without help of
recommender system cold items can be captured by innovators. An extensive experiment is
conducted on the dataset provided by Alibaba group in real-time environment to achieve accuracy
and serendipity.
Zhao (2019) discussed various recommendation algorithms and challenges in terms of big data
solutions. Author has proposed a scalable and distributed recommendation system based on
Hadoop. Combining the properties of scalability of Map-Reduce and hybrid recommendation
system author has brought a solution to solve the information overload problem in Big E-
Commerce. In this literature author has implemented map reduce technique with hybrid
optimization algorithm to fetch the most appropriate results from the datasets.

Various literature and websites are focused on Different types of recommendation systems. It is
very helpful in I-Commerce to implement recommender system for increasing the productivity of
the business. So, it is most important to implement recommender system which will give maximum
productivity with high relevancy, accuracy, efficiency and precision.

2.3 Optimization Techniques in I-Commerce

Optimization is nothing but extracting the most relevant output from the given sets. It is very
important to carry out the researches for making the best or most effective out of given resources.
In I-Commerce it is most important to design methods which can give most relevant results. Many
optimization algorithms exist such as Ant colony optimization, Particle swarm optimization,
Artificial Bee colony optimization, Firefly algorithm, Grey wolf optimization, Blue whale
optimization, Bat algorithm, Flower pollination, Spider monkey optimization algorithm, etc. So it
is most important to identify the most relevant and efficient algorithm to fetch the most relevant
output in I-Commerce.

According to Akay (2012) Artificial Bee Colony (ABC) algorithm is an optimization technique
that simulates the foraging behavior of honey bees and has been successfully applied to various
practical problems. A set of honey bees, called swarm, can successfully accomplish tasks through
social cooperation. In the ABC algorithm, there are three types of bees: employed bees, onlooker
bees, and scout bees. The employed bees search food around the food source in their memory;
meanwhile they share the information of these food sources to the onlooker bees. The onlooker
bees tend to select good food sources from those found by the employed bees. The food source
that has higher quality (fitness) will have a large chance to be selected by the onlooker bees than
the one of lower quality. The scout bees are translated from a few employed bees, which abandon
their food sources and search new ones. In this paper, we search swarm intelligence with the help
of honey bee or the other insects. Artificial bee colony optimization algorithm is used to solve the
real parameter problems in algorithms to solve the optimization issue.

Business Intelligence tools and systems with fuzzy logic systems are used by online businesses for
the future of E-Commerce. With BI 3 and fuzzy logic, the business has more possibilities to choose
types and structures of the required information to serve customers. The factors such as usability,
familiarity and conducting business with parties with trust issues can be solved with the help of
fuzzy logic and BI. A properly designed system is an interactive software-based system intended
to help consumers to gauge the level of the risk factors in given website. Akhtar (2016) studied
useful information from raw data, documents, personal knowledge to create rules to identify the
potential risk factors in respect of conducting online transactions. Fuzzy logic and BI are capable
of processing the consumer's information. The recommendations and suggestions allowed
proposed systems would assist the management to address the need of the customers, ability to act
on market drivers, optimizations of business and customer retention.

In the world of the internet, you can’t stay away from it. If you want to make your business popular
or increase it, then you must take help of the Internet. On the tip of finger if you will get the things
which as per users choice and free home delivery, Whenever your online and you can use the
internet for the growth of business on the internet. Akhtar (2014) emphasized on the behavior of
the internet, business extension, and promotions. We can select the group of active customers who
are interested in buying the option and can approach them using IE, ID, and II. Applying this
method we can capture the best customer base we want and we can easily grow our business. In
this paper author has compared ACO and PSO to find accuracy and efficiency of the customer.
Andreou (2002) emphasized the evolving area of E-Commerce where user can interact with service
providers through a mobile and wireless network, using mobile devices for information retrieval
and transaction processing. M-Commerce services and application can be adopted through
different wireless and mobile networks, with the aid of several mobile devices. Important Factor
in designing M-Commerce services and applications is the identification of mobile users. It is
classified into two major classes: directory and transaction-oriented services and applications. The

3
Business Intelligence
mobile users need and required the classification of the m-commerce services and application and
current technologies for mobile and wireless computing and their constraints.

Bucki (2012) used the logistic structure for adequate control approach. Heuristic algorithms are
used for meeting the set criterion. The criteria to maximize the production output or minimize the
lost flow capacity of the logistic system or to minimize the tool replacement criterion are
introduced. Equations of state are given in order to represent the flow of material through the
logistic system. The main goal is to fulfill the task by set criterion. Heuristics algorithms enable to
choose the satisfactory productions sequence on the basis of which a certain element of the order
vector is determined to realize which are provided to customers by means of simulations. The
advantages are that it outweighs the costs of building a simulator in a very short period of time
proving its market value, which will increase the production unit.

E-Commerce websites are increases day by day, sometimes customer got confused that what we
should buy. Though this kind of confusion is there must be some kind of solution and it helps
customers to choose their products and way of entertainment. Charoensiriwath (2003) helps to
learn personal user interest. So the next time we can directly enter into the area which we like.
Genetic algorithm and personal are the methods which can be used to find out the area of the
interest of the person. It will help the customer or the entertainer to choose the data or thing which
he wants exactly. Time is the biggest concern in the recent era. We can reduce the time by which
he can easily search the products which we want to buy. This is the perfect recommender system.

Chu (2017) explained recommender system or recommendation system (sometimes replacing the
system with a synonym, such as a platform or mechanism). A recommendation system is a subclass
of an information filtering system that seeks to predict the rating or preference that a user will give
an element. They are mainly used in commercial applications. Recommendation systems are used
in a wide variety of areas and are most often considered playlist generators for video and music
services such as Netflix, YouTube and Spotify product recommendations for services such as
Amazon, or content recommendations for network platforms. Social networks like Facebook and
Twitter. System is also one of them. Word2vec this system is adopted in the novel recommender
system. These systems can operate using a single recording, such as music, or multiple recordings
within and between platforms, such as news, books and search queries.
Now a day’s one of the most accessible markets is the E Markets such as amazon, Flipkart, etc. As
E-Commerce is increasing day by day, there are various methods are developed for the purpose of
use of E-Commerce and its study. Ant colony algorithm which is based on the tourist route is also
one of them. Through the research, it is observed that in last some years the research on ant colony
algorithm is mainly used in various problems of TSP. Ding (2019) studied E-Commerce
environment by using the “Branching ant colony With Dynamic Perturbation to understand the
dynamic disturbance of the cities for the selected cities. This research shows that algorithm
improves demerits of long research time and easily accessible to the local one. This method tourist
can easily found their E-Commerce more effectively.

According to Evangeline (2014) E-Commerce becomes integral part of life now a days. In E-
Commerce we can also perform payments, online transactions. It creates huge opportunity for the
tourist in E-Commerce. For the tourist purpose the monopoly of the particular person, institution
had been ended now they have various choices and options available so the huge competition in
the market takes place. All the things can be easily accessible as per their demands. For tourist the
most important thing is to choose the route and make sure it will give more joy in less time. So
selecting the proper route is not easy. So using this system they can easily choose the path which
they want. This will also help in their recommendation system. This is the era of the starting of
electronic commerce in tourism. So we have to combine our new and also the old culture and
heritage so that we can easily make tourism more friendly.

Feng (2019) studied about the ant colony algorithm in rural E-Commerce logistics mode to
consider the third party distribution object. So logistic method vegetables and the fruits such
farming products are the key researched products. To utilize the problem of the third party and
easily analyzed it. Colony algorithm can be calculated using different software's. Author has also
calculated the cost, utility and efficiency of the colony. At last when the final result comes out we
can easily state or mention the difference between urban and rural and we can easily solve it. As a
result author found that this method improves and assures distribution efficiency and reduces the
logistic cost. Rural e-commerce will help farmers to connect with the world so they can easily
connect to the world. This method will help to the rural peoples who wanted to connect with the
globe and easily give solution regarding rural e-commerce problems.
The vast amount of data available on the Internet has led to the development of recommendation
systems. Georgiadis (2005) proposes the use of soft computing techniques to develop
recommendation systems. It addresses the limitations of current algorithms used to implement
recommendation systems, evaluation of experimental results and conclusion. Given that the
current share of corporate retail is less than 10% and that the national venture capital industry is
still in its infancy, policymakers should look for an approach with investments and the experience
of foreign participants. E-commerce should be seen in the context of India's profile, which should
appear in the next decade. Thus, he deserves to be considered on his dignity, and not on physical
reserves.

According to Goswami (2018) learning to rank LTR algorithm trained with behavioral features
from E-Commerce customer interactions do not by themselves address the discoverability
problem. Instead, an explore-exploit framework must be integrated with the ranking algorithm.
Strategy based on monotonic sub modular optimization function that needs to be integrated with
linear LTR models also works well for discovery and has nice performances with respect to sales,
revenue and relevance. Emerging E-Commerce search problem is an optimization problem with
multiple objectives including sales, and discoverability besides relevance. The theoretical
framework could enable many interesting ways to further formalize the E-Commerce search
problem and develop a new effective E-Commerce search algorithm.

According to Guo (2018) modern computers are powerful enough to support multi-component
software that manages applications in electronic commerce and communication networks are now
fast enough to send/receive large amounts of information needed between computers that
communicate almost in real time. However, this does not apply to mobile commerce applications.
Although mobile commerce is based on the same basic information technologies, such as the
internet and the World Wide Web, very different infrastructure components are at the core. The
efficiency of M-Commerce web sites in supporting users with the content they need in the most
optimized approach is considered as the way of holding them.

Humphreys (2016) explained the basic particle swarm optimization algorithm (PSO) technique
consists of two code lines and to use these two code lines, the problem of optimal slave control
must be solved in order to provide a quick and autonomous solution or result from a hybrid
optimization technique. Disclosure of this hybrid optimization technique using the PSO algorithm
with a two-dimensional model in a static threat atmosphere guarantees the use of this method to
continue future research. Also, this technique may be employed for 3-dimensional model for
solving problems include stochastic threats and capacity to dynamically re-plan the task in the
middle of varying mission atmosphere like a pop-up threat.

More options more confusion, the era of E-Commerce confused peoples many times. So user has
some filters by which they can directly connect to the material which they want. It is most effective
in cases when there is in-band noise present. In those cases AVT is better at filtering data
then, band-pass filter or any digital filtering based on a variation of. Conventional filtering is useful
when signal/data has a different frequency than noise and signal/data is separated or filtered by
frequency discrimination of noise. Those filters are created using passive and active components
and sometimes are implemented using software algorithms based on FFT. Jiang (2018)
implemented AVT filtering in software and its inner working is based on a statistical analysis of
raw data.

Khera (2008) discussed the lack of semantic factor in recommendation systems and describes the
different recommendation techniques that are being employed in the current e-
commerce marketing. Recommendation system can be broadly classified into three categories:
content-based, collaborative, and hybrid recommendation approaches. Content-based systems
consider the properties of the items to be recommended. This developing India will consist of
consumers who will have the desire, need and conviction to use the Internet for a number of
reasons, among which there will be one purchase. E-commerce is often combined with corporate
retail. Therefore, it deserves to be considered on its own merit, instead of being clubbed with brick
& mortar retail.

The main objective of this method is used to turn passive visitors into active users in order to
improve a website’s performance. The results of both A/B tests showed an increase in conversion
with design. Kaushik (2017) evaluated the results and its statistical significance was done using
both optimized statistical engine and null hypothesis testing. Metrics such as high exit-rates
combined with many page view and high revenue-generating pages will allow e-merchants to
identify where to start their optimization process. There are different approaches to evaluating the
statistical significance of A/B tests. The selection of the method depends on what data and subject
area one is to collect and examine. It helps to increase in sales, revenues, and profit also reduction
of customer acquisition costs.

Mohammad (2015) emphasized on the rapid development of mobile telephony has stimulated the
growth of mobile Internet as the basis of mobile e-commerce. E-commerce applications tend to go
beyond the organization and are able to adapt to changes both in the competitive market and in
technology in general. But typical mobile commerce client devices, such as cell phones, have
limited resources (such as screen size, memory, etc.). The customization technology is focused on
making the site more responsive to the individual needs of each user. Therefore, personalization
can be used as a filtering mechanism that delivers information that mobile devices can efficiently
process.

Nasiri (2018) explained whale optimization algorithm approach for clustering nothing but values
of attributes by using them we can identify homogeneous group which will be helpful in the mining
data which is the important part of the clustering. In computer science and mathematical
optimization, a metaheuristic is a higher-level procedure or heuristic designed to find, generate, or
select a heuristic (partial search algorithm) that may provide a sufficiently good solution to
an optimization problem, especially with incomplete or imperfect information or limited
computation capacity. Metaheuristics sample a set of solutions which is too large to be completely
sampled.

Metaheuristics may make few assumptions about the optimization problem being solved, and so
they may be usable for a variety of problems. Nayak (2013) compared different optimization
algorithms and iterative methods but metaheuristic do not guarantee that a globally optimal
solution can be found in some class of problems. Many metaheuristics implement some form
of stochastic optimization so that the solution found is dependent on the set of random
variables generated. In combinatorial optimization, by searching over a large set of feasible
solutions, metaheuristics can often find good solutions with less computational effort than
optimization algorithms, iterative methods or simple heuristics. As such, they are useful
approaches for optimization problems. Several books and survey papers have been published on
the subject.
Niknam (2009) explained variety of mobile commerce applications and a variety of goals and
customization methods imply that none of the existing methods can satisfy all needs. However,
recommendation systems are considered an integral part of personalization because they include
the experience of the user community. Personalization can be defined as user interface attributes
that are computer-driven, often by an algorithm or user model based on the user’s selections,
habits, demographic information etc. To support personalization mechanisms, M-Commerce user
interface of applications is concentrated on the individual requisites of the persons in contact with
it.

An important part of the tuning functions is implemented by recommendation systems. These

modules provide predictions, opinions, and recommendations that help the user evaluate the
elements. Referrers can be a valuable part of mobile browsers, as they can significantly reduce the
amount of information a user must look for in order to find elements of interest. Panigrahi (2016)
proposed a method to a wide range of online business activities for products and services.

According to Papamichail (2007) in this digitalized era, the market is flooded with various online
marketing tools and for 2 years the effect of E-Commerce is phenomenal. It includes the use of
electronic commerce with an analysis of the use and impact of electronic commerce in India's
textile industry. The benefits of E-Commerce include global marketing opportunities for products
and markets. The main advantages of E-Commerce are global marketing opportunities for products
and markets. However, a key limitation of E-Commerce is the risk of channel conflict between
existing customers and new profile customers. The current growth in e-commerce in the textile
industry was driven by the emergence of new companies and the support of venture capital and
entrepreneurship. Currently, the Indian

According to Robert (2011) supply chain is one of the key areas in e-commerce. The supply chain
is generally defined as multi-system of operators, manufacturers, distributors, resellers and
customers, among which is a flow of goods, information and finance. The quality of supply chain
largely depends on the set management processes. SCM concept includes not only the logistics
process as well as strategic management of entire supply chain including supplier selection,
distribution of production functions and outsourcing of capacity or processing of customer
requirements. The aim of the supply chain is to deliver goods in shortest time at the lowest possible
price while ensuring the highest quality.
We collect huge data so it is also important to make a grouping of that data. Rana (2010) explained
different data mining techniques. Computer cluster is a technique of linking many computers
together to act as a single computer. Data cluster allocation of contiguous storage in databases and
file systems. Cluster analysis is a set of machine learning algorithms to group multi-dimensional
data-set into closely related groups (such as KNN algorithm) in hash tables, the mapping of keys
to nearby slots. Jasola (2010) Combined together Data clustering and PSO application are used
widely to grouping and data mining. This will gives you total differences between algorithms such
as GA and SA. It helped in optimization and gives you proper direction, efficiency, environment.
According to Santucci (2011) optimization is a common problem in a business. In computational
science, particle swarm optimization (PSO) is a computational method that optimizes a problem
by iteratively trying to improve a suitable solution with respect to a given quality measure. Solve
the problem by having a set of possible solutions, curved particles and moving these particles in
the search space, according to simple mathematical formulae over the particle's position
and velocity. Each particle's movement is influenced by its local best known position but is also
guided toward the best-known positions in the search-space, which are updated as better positions
are found by other particles. This is expected to move the swarm toward the best solutions.

Sethi (2017) implemented a method to recommend the proper products to the customers and helps
customers make the decision during the purchasing process. The recommender system could be
operated through the top sellers on the website, the demographics of customers or the consumers'
buying behavior. However, there are 3 ways of recommendations: recommending products to
customers directly, providing detailed products' information and showing other buyers' opinions
or critiques. It is beneficial for consumer to experience without physical shopping. In general,
recommender system is used to contact customers online and assist finding the right products they
want effectively and directly. Author has emphasized on the filtering of the recommended system
& it also compares two algorithms. Stay updated that's all for the era of E-Commerce which is
done by this method it saves time too.

Soonsawad (2013) discussed conversion rate optimization in online-retail environments. It

examines how one company has created its own model and used model to successfully enhance
internet business. It identifies the variable that influences online customer behavior and discusses
the importance each of seven elements which are catalyst, persuasion, usability, interactivity, trust,
aesthetics and marketing mix. It identifies two additional key elements in conversion rate
optimization and develops a new model that incorporates the two new elements. This new
hierarchical model presents the elements stage-by-stage to help online retailers to convert visitors
into customer. The new conversion rate optimization framework provides organizations with a
practical framework that takes into consideration the customers level of web experience and their
decision-making process. The framework helps to develop websites that appeal to customer and
convert more visitors into buyers. A new model has been created to assist managers and website
developers to create sites that will improve conversion rates. In addition, the effect of product type
and consumers experience at website should be also considered in investigating consumer
behavior.

The rapid growth of E-Commerce websites also makes us to think what should buy, which should
buy and where should buy? So there might be a search engine which gives us a details comparison
between the two E-Commerce websites. Not only clearing the doubt but also it must understand
customer psychology i.e. which product is cheaper, long lasting, Easy, user friendly by doing this
all filters we can compare E-Commerce websites as its growth day by day. By doing this Shao
(2010) studied a how one can easily expand his business online and touch the sky and reflects how
E-Commerce Comparison shipping model of neural network based on ant colony optimization
help us.

Sunda (2017) emphasized on techniques such as directory submission, keyword generation and
link exchanges. This technique can be applied as a marketing technique and to get top listing in
major search engines like Google, Yahoo and MSN. E-commerce companies can grow their sales
and revenue using SEO4. It consists of an introduction to e-commerce search engine optimization,
search engine algorithms, ranking factors and efficient techniques to optimize an e-commerce
portal. SEO is the technique of getting organic traffic through search engines and that is done by
optimizing web pages or entire sites in order to make them search engine friendly. SEO simply
translates your website or blog into language that is easily understood by search engines, Hence
when user search for the offer, product, services, information or answers to specific questions. It
improves the volume and quality of traffic to a website from search engines.

Buy one get two is what we are looking today. For attracting the customer's e-commerce
companies are giving various offers or schemes. Human tendency is that we shouldn't think much
more if we get things in cheaper amount whether it is cheap quality, healthy, useful without
thinking anything sometimes we buy it. Sukstrienwong (2011) applied some algorithms to find out
the maximum profit. It contains the huge group of buyer, so we can decrease the value and got in
big discount. The result which we got from this method i.e. ACO-FCB is compared to the genetic
algorithm. If we buy a bunch of the products it will cost us in a less amount if we make a sum of

4
Search Engine Optimization
individual product. There is no limitation to the seller he can sell as he can or wanted to sell i.e.
freedom of working.

Xu (2014) explained about the most important factor about E-Commerce websites is the
recommended system. The contemporary E-Commerce trend recommends companies to shift the
traditional business model where focus on "standardized products, homogeneous market and long
product life cycle" to the new business model where focus on "varied and customized products".
E-Commerce requires the company to have the ability to satisfy multiple needs of different
customers and provide them with wider range of products. With more choices of products, the
information of products for customers to select and meet their needs becomes crucial. In order to
address the mass customization principle to the company, the use of recommender system is
suggested.

When you think about your business, it must be planned and optimized. It is just to achieve the
maximum profit in minimum time span. For that we need very few resources but the right direction.
It tends to improve the quality of your products and services. To achieve this will deal with PSO.
Yang (2012) described usage of PSO and non-linear machine learning algorithm to find difference
between them. Accelerated Particle Swarm Optimization, Support Vector Machine or
Metaheuristic Support Vector Machine with APSO is the techniques used to find the optimized
data.

Optimization techniques can be used to reach the optimal solutions. As per the above literature
survey it has been observed optimization techniques may help to achieve the most relevant or
optimal results for the I-Commerce. Many techniques and researches are conducted for
recommending the relevant results using optimization techniques. But the accuracy of this result
is very less. Also, it is very hard to predict the most appropriate products for user. This study gave
me overall glimpse of optimization techniques and helps to select a appropriate optimization
algorithm.
Different Naturally inspired Optimization Algorithms are [3]

1. Hybrid glow worm

2. Salp swarm algorithm
3. Teaching–learning-based optimization
4. Firefly algorithm
5. Water cycle algorithm
6. Cuckoo search
7. Imperialst competitive algorithm
8. Swarm optimization algorithm
9. Mine bomb algorithm
10. Whale optimization algorithm
11. Biogeography-based optimization
12. Bacterial foraging optimization
13. Quantum-behaved particle swarm optimization
14. Design by shopping paradigm
15. Dolphin echolocation algorithm
16. Artificial fish swarm
17. Flower pollination algorithm
18. Artificial flora optimization algorithm
19. Whale optimization algorithm
20. Cockroach swarm optimization
21. Grey wolf optimizer
22. Invasive weed optimization
23. Albatross wind shear algorithm
24. Cat swarm optimization
25. Moth-flame optimization algorithm
26. Galactic swarm optimization
27. Hummingbirds optimization algorithm
28. Bird mating optimizer
29. Artificial bee colony algorithm
30. Fruit fly optimization algorithm
31. Big bang - big crunch algorithm
32. Harmony search algorithm
33. Ant colony optimization
34. Bat algorithm
35. Penguins search optimization algorithm\
2.4 Big Data in I-Commerce
Big data is a large amount of data. Due to the continuous increase in velocity of data, size of data
is increasing day by day. In optimization, machine learning algorithms it is most important to
process a large amount of data to predict most relevant output. Big data deals with 5 V’s i.e.
Velocity, Variety, Volume, veracity and value. I-Commerce deals with unstructured data. So it is
most important to process large amount of data, we need a high-end systems for processing this
data.

According to Akter (2016) in recent times big data analytics play a huge role in E-Commerce
domain. But it is not explored as much and this restricts its theoretical and practical development.
In this paper, authors ought to explore the definitional aspects, distinctive characteristics, types,
business value and challenges of big data analytics in the e-commerce landscape. Discussions also
include future research challenges and opportunity in theory and practice. Big data with its unique
characteristics i.e. voluminous, variety, velocity and veracity can be easily distinguished from
traditional form of data. Big data has emerged as an innovation in E-commerce due to challenges
created by information evolution. Many of the big firms have already embraced big data analytics.
Once it is well understood it will maximize the business value.

In today's technological world, personal data has accumulated to a point where it no longer fit into
the memory of the computer. Big data offers ability to extract information to produce useful
insights into significant value. The digital data is exponentially being created, replicated and
consumed. This growth is attributed in part of information sensing technologies propagation. This
volume of data created is of economic value for marketers, as they seek effective way to benefit
from the big data. According to Eastin (2016) consumer's privacy is being risked in rapidly
increasing big data ecosystem. Consumer increasingly uses mobile technology can be benefited to
the users by showing personalized mobile advertising. Thus scholars are challenged to create
framework that considers both preventions of risk as well as motivates consumers to continue
sharing their personal information.
Kwon (2014) emphasizes on improving the capabilities of a firm and its performance. Most of the
leading companies are actively adopting big data analysis to build their market competition and
open up new business competition. But many firms are new to this adoption curve due to lack of
understanding of and experience with big data. Hence, a research model is proposed to explain the
acquisition intention of big data analytics mainly from theoretical perspectives of data quality
management and data usage experience. Empirical investigation reveals that a firm's intention for
big data analytics can be positively affected by its competence in maintaining the quality of
corporate data.

Big data analysis is also helping E-Commerce in terms of logistics management also. Evolution of
e-commerce has complicated its logistics, management. Lin (2018) has proposed model
incorporates theories of resource dependence theory and innovation diffusion theory to generate
logistical strategies. Ten critical themes of e-commerce logistics from topic mining are “E-
commerce order management”, “Europe e-commerce trends”, “India’s e-commerce logistics”, “E-
commerce order management”, “E-commerce logistics cloud services”, “E-commerce
distribution .2 management”, “Tax policies”, “E-commerce logistics platforms”, and “E-
commerce logistics networks”. This proposed hybrid content analytics framework provides a
research foundation for e-commerce logistics management.

Bhatia (2015) explained the importance of big data in E-commerce and offers a wide range of
product in E-commerce websites. In order to remain competitive the firms formulate marketing
strategies based on real time data. These E-commerce firms are finding ways to get information
from big data to effectively communicate with consumers to provide better service. These various
functions of firm involve personalized service, competitive pricing, fraud detection and handling.
Different sources of big data in E-commerce involve data generated from real time sensors, social
media, tweets and other business transaction information. E-commerce analyses cookies to identify
patterns in customer habit and accordingly satisfying consumer's expectation. Big data enables to
understand customer better which in turn leads to meeting customer need.
To process large amount of data either we need to use high-end services or distributed architecture
or Hadoop. Where we can perform large tasks with reduced time factors and increased efficiency.
So we need to use some distributed architectures with parallel clustering algorithms. It helps to
reduce the time factor and increase efficiency.

2.5 Parallel Clustering Algorithms in I-Commerce

Parallel clustering algorithms are used to process the data in parallel way with minimum time and
high efficient output. Many parallel clustering algorithms such as K-Means, Fuzzy C-Means are
used to process the large amount of data in minimum time. Parallel clustering algorithm has very
high efficiency and strong elasticity when dealing with a huge number of data items but on the
other side, clustering effects of this algorithm are affected by clustering number and initial
positions of clustering center.

Huang (2014) implemented a system for dealing with the large number of high-dimensional
transaction data, clustering algorithms consist the shortcomings like bad elasticity, weak handling
ability of high-dimensional data, sensitive to time sequence of data, bad independence of
parameters and weak handling ability of noise. When analyzing e-commerce transaction, the K-
Means clustering algorithm shows high dissimilarity and inter-class similarity compared with that
of the Single-pass clustering algorithm.

In case of High transactional data clustering algorithms are face various problems like bad
elasticity, weak handling data, high dimensional data and various noises in clusters. For
experimental purpose Humphreys (2014) had taken a sample of 300 mobile phones. After
performing operations he has concluded that K-Means algorithm has inter class similarity and high
interclass dissimilarity while analysing E-Commerce transaction. It is also observed K-means
algorithm has high efficiency and strong elasticity while dealing with large amount of data. It is
affected by clustering center and initial position.

Kim (2009) explained the importance of Iterative and recursive procedures are required for many
clustering algorithms so this task becomes computationally expensive. Hence parallelization of
clustering algorithms is unavoidable. Clustering algorithms are classified into Partitioning,
Hierarchical, and Evolutionary approach, Dense-based, Model-based, and Graph-based. Among
these types, Partitioning and Hierarchical algorithms are most popular algorithms. Parallelism in
these algorithms is used for efficient clustering strategies and distance computation. K-means
algorithm represents a partitioning clustering algorithm.

Clustering algorithms are plays vital role for the defence services in each country. It can be used
in unsupervised learning, exploratory pattern analysis and image based applications. Many
clustering algorithms like partitioned or hierarchical clustering algorithms require intensive
computations. In this paper author used clustering algorithms with 2 n patterns and 2m features. It
has been observed time complexity of traditional systems with single processor computer is O
(MNK). Li (1989) proposed a SIMD with MN processors to reduce the time complexity.

Li (2007) proposed a new algorithm for message passing multiprocessor subsystem. In this
algorithm we have combined prediction concept with bisector K-Means algorithm. Bisector K-
Means tends to create similar clusters with smaller entropy. Parallel bisecting K-means with
prediction exploits data parallelism and adopt prediction step to manage the workloads of multiple
processors to achieve high speed. Experimental results shows task is completed with high speed
and reduced time.

Papamichail (2007) describes the K-means range algorithm. It is a combination of spatial data
structure with partial K-means clustering algorithm. It is named as range tree, which allows the
searches in fast range. It is use to find out the real time solution for development in E-Commerce.
It allows customer to select a products in multiple dimensions i.e. various range of products with
variety of choices and search for product information. This information can be used to enhance the
purchase decisions. This literature also explained the advantages and implications of this
development in shopping environments. This approach can work in E-Commerce and M-
Commerce environment.

Thomas (2011) proposed a method and select clusters from the large data repositories using K-
means algorithms while QoS is also taken into account. The Map-Reduce technique is used to
parallelize the K-means clustering algorithm, further improving computational efficiency. “Hence,
parallel K-Means algorithm is explored for finding the best effective cluster containing the hops
which lie in the best cluster with the best throughput in self-aware MANET5.” By interacting with
the layers of the protocol stack, there is an attempt of the addition of the self-awareness in the

5
Mobile Ad-hoc Network
MANET to find the status of the QoS parameters. This parallelized clustering algorithm enhances
the performance of the nodes connected in MANET along with meeting the QoS constraints given
the suitable path is selected between the nodes in MANET.

Wijaya (2017) studied the ecology of “Henz Collection” shop which sells clothes, shoes and bags.
The problem with shop is they don’t have a website which will sell the products as per the loyalty
of customers to shop online. To sort out this issue Fuzzy C Means algorithm is used to predict the
discounting seen. Every data point is considered as a degree of membership. Value of centroid is
taken randomly and centroid determination is determined the output is not changed. This method
is generally used to avail the rebate/discount to the customers. It has been found that in 2014 degree
of similarity is 60% is increased to 80% in 2015.

In intelligent traffic system traffic subarea division is most important for traffic management
system. Many existing methods can works with fewer amounts of data. If Data is more they are
not efficient to give efficient output. To solve this problem Xia (2015) implemented par3PKM
algorithm. It is a combination of map reduce algorithm with 3 phase K-Means algorithm. In this
algorithm firstly K-means algorithm is applied in highly distributed environment using Hadoop to
perform Map-Reduce activities. Experimental evaluation result gives more accuracy, higher
scalability high efficiency and can divide the traffic subdivisions effectively.

Xiaowe (1999) implemented a DBSCAN algorithm which relies on density based cluster and
indexed of to discover cluster with arbitrary shape and distinguished noise. Here author has
presented a parallel version of the algorithm called as PDBSCAN. Here shared nothing
architecture is used with multiple computers. It uses a distributed data structure. DR∗-tree with
spatial distributed data structure is used to spread the data among the multiple computers. Index
of data is replicated on multiple computers with distributed spatial index. A performance
evaluation is shows that process is linearly speedup and has excellent scale up behavior.

To solve the various E-Commerce problems like massive data processing, data sparseness, fast
clustering of different types of data. Xie (2012) proposed a novel methodology using parallel
clustering based methodology based on FP tree. In this methodology algorithm is improves by two
ways. First is usage of Map Reduce algorithm to implement X-means algorithm to perform the
parallel clustering methodology. Second is FP tree structure usage for solving Boolean data,
collection of type’s problem which has no definition about mean and usage of K-Means clustering
algorithm. This is a parallel clustering based algorithm with FP tree. It is also called as PXM
algorithm.

Zhang (2013) explained a method for exploratory analysis for statistics, image processing and
pattern recognition. Here K-means algorithm is used because of its simpler implementation,
efficiency and empirical output. Real world applications are produces huge amount of data hence
the efficiency of mining data tasks are challenging and significant issues. To solve this problem
author has proposed a MPI with K-means model. MPI is a message passing interface advantage
with scalability, efficiency and performances. In the parallel environment it enables working of
clustering algorithm effectively. Experimental study of MK means is stable and portable and it
gives efficient output with reduces time on a huge volume of data.

Above literature survey clearly stated that parallel clustering algorithm can be used to complete
the task to reduce time with high speed and efficiency. Many parallel clustering techniques have
been evolved. K-Means Fuzzy C means, Map-reduce etc. All these algorithms give better results
in various environments.

2.6 Prediction based Hybrid Optimization technique for I-Commerce

It is very significant to conduct research in the area of I-Commerce. As I-Commerce is most
trending field and till 2022 whole world is going to turn out business through online ways. So it is
very important to develop a process where we can conserve customer satisfaction and gives the
proper services to customer. It has been observed many researchers have done the various
researches in the arena of E-Commerce or M-Commerce. They have used different optimization
algorithms or Machine learning algorithms to predict the best product or recommend good
products to the customer.

Avasthi (2013) carried out the prediction of mobile users’ behavior with the help of clustering
algorithms. Temporal property and user relations were employed in this paper. The CTMSP–mine
technique was employed to extract the users’ behaviors. CO-Smart-CAST has created the user
clusters as well as similarities among the mobile users. Here, the particular time intervals were
used to group the big mobile logs, which are identified with the help of genetic algorithm-based
technique called Get Number of Time Segmenting Points. The CTMSP-mine algorithm has used
both the results of time intervals and CO-Smart-CAST. These patterns were employed to forecast
the future behavior of mobile users and then service recommendations.

Cavazos (2006) designed new class of optimization heuristics is known as the hybrid optimization
method. The implementation of hybrid optimization techniques has examined many optimization
algorithms. The hybrid optimization technique uses heuristic optimization to select the best
algorithms among all that will be used in certain conditions. This method will minimize
compilation efforts and use an efficient algorithm in many cases, but it rarely uses the more
efficient and expensive optimization algorithm when evaluating the greatest benefits that efforts
can measure. This transaction or compensation is very attractive if the compilation time is
significant, for example, a just-in-time compilation environment.

Dong (2016) describes the extension of the approach to recommendation that combined a simple
overlap-based model of feature similarity with sentiment analysis. This work described an
improved approach to combining product similarity and feature sentiment which provided much
greater configuration flexibility and accommodated a wider range of similarity assessment
techniques. Second, the method described two different approaches to computing feature sentiment
that avoids the ad hoc treatment of so-called residual features. Hence, use of clustering techniques
as a way to identify common features and that might be treated as independent features using
standard opinion mining approach. The limitation of this work relates to the choice of techniques
which were adopted for the purpose of opinion mining and sentiment analysis.
Evangeline (2013) presented the analysis on service pattern prediction and mining the personal
mobile commerce environment using 2-DML association rule mining technique. This research has
performed theoretical analysis of the mining as well as prediction of patterns in the mobile
commerce environment. Forecasting the behavior of mobile user is the most significant concern;
in order to predict the mobile users’ behavior frequently in efficient manner then the pattern mining
technique was employed. The benefits of data mining technique include retailing, banking, law
enforcement, etc. In this research, the 2-DML association rule mining technique was employed to
determine the location based service patterns in the mobile commerce. The result of this algorithm
has revealed the best performance efficiency as well as memory efficiency.

Evangeline (2014) forecasted the frequent pattern mining in the mobile commerce atmosphere
using systolic tree based technique. In this research, a new mobile commerce framework was
employed for mining as well as predicting the behavior of mobile users in their movements and
tradeoff. It was used to suggest the products and stores to the new user. When mobile customers
travel among the stores, then the system forecasts the repeatedly moved products to the user
depends upon their choice that contains mobile information like user identity, products purchased
and stores were saved in the mobile tradeoff database. The new Systolic Tree Based technique is
an effective and efficient algorithm to extract the repeated item sets. It has also formed best
performance results when compared to other algorithms.

Humphreys (2016) analyzed about the hybrid optimization technique that adopted to control the
optimization problems. The Direct Orthogonal Collocation technique was the fast technique to
solve the optimization problems, but the challenge was picked for the gradient based Non-Linear
Programming, which converse quickly without receiving any ‘stuck’ in the undesirable optimal
region. Heuristic techniques have provided the opposite challenge; they are very hasty at
determining the global area but they can’t exposure to the worldwide optimal solution.

Kandasamy (2015) predicted the behavior of mobile users in the location based services using
MPSO-OTI26 technique. The mobile user behavior prediction is a significant concern. The MPSO 7
technique was aimed to determine the optimal time interval in forecasting the behavior of mobile
users for the location based services. Also, this MPSO technique was targeted to find the better

6
Modified Particle Swarm Optimization Based Optimal Time Interval Identification
7
Modified Particle Swarm Optimization
time interval and low computation complexity. The segmentation of time has been performed by
MPSO algorithm. MPSO-OTI2-PMB has provided high effectiveness (recall, precision and F-
measures) for predicting the behavior of mobile users in the location based services.

Kim (2009) proposed data properties based on the various kinds of clustering model includes
hierarchal, partition, graph based clustering, dense-based, model-based and so on. Many of the
parallel clustering algorithms are employed as the message passing model in the master slave
architecture and it may be used for effective computing. The general application of parallel
clustering algorithms is image processing. Parallel bottom-up clustering algorithms are used to
partition the circuits in VLSI, unlike many top-down clustering algorithms for the layouts, the
layout of bottom-up clustering algorithms are helpful in increasing the size of circuits.

Lee (2011) predicted the hybrid techniques of genetic algorithms and neural networks for the
controls of internet based systems for business to customer e-commerce. The main intention of
this research was to propose ISB2C-NNGA technique, a hybrid optimization technique with
genetic algorithms and neural networks for the framework of ISB2C 8 controls that uses BPN9
technique as the prediction of controls by electronic commerce environment; and GA 10 was used
as the pattern search tool to evaluate the exponent of the independent variables in the multivariate
regression model. The cause of each control modes on the implementation may be determined with
the help of standardized coefficients and exponents in GA based non-linear regression analysis in
the internet-based system.

Maalini (2014) performed the prediction of mobile users’ behavior in the location-based service
atmosphere. In this research, a new Cluster-based Temporal Mobile Sequential Pattern Mine
algorithm was employed to realize the sequential patterns. A prediction strategy was proposed to
estimate the consequent mobile user behaviors. In CTMSP- Mine11 algorithm, user clusters were
built through a new algorithm called Cluster Object based Smart Cluster Affinity Search algorithm
and determine the similarities among the customers using location based service alignment
technique. The time segmentation technique was provided in order to identify the time intervals in

8
Internet-based System for Business-to-Consumer
9
Back Propagation Neural Networks
10
Genetic Algorithm
11
Cluster-based Temporal Mobile Sequential Pattern Mine
which similar mobile features exist. By the experimental evaluation under different simulated
conditions, the CTMSP-mine techniques have shown to deliver the best performance.

Nilashi (2015) implemented a new recommendation model to improve the recommendation quality
and predictive accuracy of Multi-Criteria Collaborative Filtering The main objective was to solve
the scalability and alleviate the sparsity problems in the MC-CF12. The experimental results
demonstrated that the combination of Neuro-Fuzzy and dimensionality reduction techniques
remarkably improved the recommendation quality and predictive accuracy of MC-CF in relation
to the previous recommendation techniques based on multi-criteria ratings. The limitation of the
methods and recommendation model were evaluated solely on one multi-criteria dataset without
incorporating tags and content of users and items.

Mitra (2016) analyzed the various properties of the graph like clustering co-efficient, degree
distributions, etc. and tried to reason out the underlying relation between such distributions. In an
attempt to understand the modification in buying pattern, the paper proposed an algorithm to mine
patterns of interest. This type of information was used for the purpose of building a prediction
model. The experimental results observed that the recommendation can be good for a large part of
the dataset but may not ensure a highly precise recommendation for every item. The disadvantage
of the approach was testing the efficiency of this recommendation system was computationally
expensive for every item.

Nayak (2013) performed the prediction and mining of location based service patterns in the
personal m-commerce using Associated Rule Mining technique. This research has several benefits
over the consumption of data mining tools include: the data has stored at the data mining, hence,
the management of privacy and security concerns were greatly decreased; there was no separate
data mining tool, instead use query processing of data mining itself; this framework has permitted
ad-hoc data mining queries on the entire data mining process and it has expanded the associated
rule mining domain from the transaction level to aggregated level data. In this research, there are
nine various features include probable data structures, intended user groups, implemented targets
and techniques, export and import abilities, interaction styles, license policies and platforms were
changeable.

12
Multi-Criteria Collaborative Filtering
Niknam (2009) the partition clustering algorithms attempt directly to decompose the dataset into
a disjoint cluster’s set. The criterion function in which the clustering algorithm attempts to reduce
the local structure of data by allocating clusters to the peaks in global structure or probability
density function. The global criteria can reduce some dissimilarity measures in samples inside each
cluster while reducing the dissimilarity of various clusters Clustering is the process of separating
input data sets known as ‘clusters’.

According to Oreski (2014), the hybrid techniques of combinatorial optimization area rising
research region that designed to solve the difficult optimization problems. By using the
perspectives of relations between the combinatorial optimization techniques, the strategies of
merging them and the ideas for solving the problems of optimization, this research has presented
an instance of hybrid method for the feature classification and selection in the credit risk
assessment. This analysis has emphasized the significance of hybridization as the cooperation
concept between meta-heuristics as well as other optimization methods. The proposed algorithm
has optimized the average prediction accuracy as a fitness function. If the banks need to optimize
a cost function instead of accuracy function, the hybrid genetic algorithm and artificial neural
networks technique maybe solve this request with the help of algorithmic methodology that is, by
changing the fitness function.

Papamichail (2007) analyzed the modified data clustering in electronic commerce using k-Means
Range algorithms. This research has employed the k-means range algorithm and a mixture of
partitional k-means clustering technique together with a familiar spatial data structure that is range
tree, which permits quick range searches. It has provided a synchronized result for the growth of
distributed interactive decision that helps in electronic commerce. Since, it permits the customers
to model their better choices along with many dimensions, explore for product information and
form the products’ data clusters that retrieved to improve their purchase decisions. This research
has also described about the implications of both customer as well as online store approach.

Priyadharshini (2013) determined an effective framework for forecasting the mobile commerce
patterns using Graph Diffusion technique. The problem of mobile commerce pattern mine
technique has been overcome through the effective framework with the help of graph diffusion
technique. The main intention of this research was to build the graph for the purchased items
through mobile users and after that to determine the repeatedly purchased items. With the help of
ranking technique, the researcher has ranked the number of items depending upon their tradeoff or
transactions. By estimating the behavior of mobile users as well as recommended the ranked items,
this framework has produced more effective and precise item recommendation when compared to
Mobile Commerce Explorer framework.

Priyadharshini (2013) predicted the mobile commerce patterns of users by Collaborative Filtering
algorithm. This research has been performed using mining and predicting the behaviors of mobile
commerce users such as movements as well as their purchase transactions. The main objective of
this research was to analyze the movements of mobile users to new places rather than considering
the frequent moving places alone. A framework of m-commerce explorer has been implemented
in order to create the recommendations for products as well as stores through analyzing the mobile
users. This research was used to recommend products and stores in new places through considering
the items’ rating that provided by other users in the new places.
Tabriz (2017) studied a mixture of similarity-based and trust-based methods that was used to
calculate users’ similarities and trusts and then predict the ranks by mixing the two methods. The
least overlapping was calculated by using the clustering technique namely, k-means method. The
parameter specified in the method was rating matrix, number of clusters and number of
recommendations for the target user, respectively. The method provided better performance while
comparing with the existing methods. The method mainly works on trust relationship between
users. Hence other users who need more information regarding items cannot able to retrieve the
data in large datasets.

Witten (2016) proposed a clustering algorithms have been employed in various fields like
medicine, manufacturing, nuclear science and radar scanning. Data clustering algorithms may be
partition or hierarchal. In hierarchal clustering algorithms, the data aren’t clustered into particular
cluster in one step. Instead, a sequence of partition scan occurs from

Xu (2017) proposed a new Co-clustering collaborative filtering model with Review Text (CMR).
Specifically, the method employed the co-clustering technique to model the user community and
item group. Each community group pair corresponds to a co-cluster, which was characterized by
a rating distribution in exponential family and a topic distribution. The paper extended CMR and
developed an enhanced model called Time-aware Co-clustering collaborative filtering Model with
Review text (TCMR) to consider time information and exploit the temporal interactions among
review texts and co-clusters of user communities and item groups. The proposed TCMR model
performed better than CMR and the standard time-aware recommendation model on the rating
score prediction tasks. The method also investigated the temporal effect on the user–item co-
clusters.

The literature review reflects many researchers have developed different optimization algorithms,
parallel clustering algorithms to predict the most relevant data in I-Commerce. Also, a researcher
has implemented hybrid techniques where they have combined to different optimization
algorithms or parallel clustering algorithm with optimization algorithm or hybrid recommendation
systems to predict the relevant products in I-Commerce.
2.7 Comparative Analysis
Sr No Topic Summary
It has been Observed I-Commerce is most trending
field in the recent era. I-Commerce is a combination
of E-Commerce, M-Commerce and IoT Commerce.
1 I-Commerce As per this survey, it is most important to conduct a
research for predicting the products in I-Commerce.
Impact: It motivates me to select I-Commerce
area for research.
Recommendation system plays an important role in
the selection of products in I-Commerce.
Recommendation system works with either
collaborative filtering or product based filtering. To
perform this activity a most impeccable process need
Recommendation
to be developed. Which will help the users to select
2 system
appropriate method; But process must be unique,
perfect and authenticable.
Impact: It helps to select the appropriate
recommendation system and also helps to
understand the different aspects of
Recommendation System.
It is well said that natural system or nature-based
technologies give better results and output of such
processes are more efficient and relevant. As per
survey it is observed as compared to machine
3 Optimization techniques learning algorithms, Output generated from Natural
techniques is most efficient. By keeping above
results it motivates to select appropriate optimization
algorithm for the selection of relevant product. Also
survey is being conducted through different
 optimization algorithms and it is observed Particle
swarm optimization algorithm and Grey wolf
optimization algorithm is most efficient as compared
to Ant Colony Optimization, Ant bee colony
optimization, blue whale optimization, money king
optimization and other optimization algorithms.
Impact: As per above survey Particle swarm
optimization or Grey wolf optimization, Machine
Learning algorithms like decision tree, random
forest, Naïve bayes algorithm can be considered
for the optimization of keywords.
Big data is a large amount of data and to process a
large amount of data we need to select appropriate
methods for the processing of large amount of data.
For our research, we are considering large amount of
data. As per survey if the method is not appropriate
4 Big Data
it will take large time to fetch relevant output or
reduces the system performance.
Impact: It motivates me to conduct research
where a process needs to be developed which will
process the big data in efficient way.
Parallel clustering techniques are used to process a
large amount of data in parallel way by reducing
time and improving efficiency. Many researches are
conducted by various researchers. Some researchers
Parallel Clustering have used map-reduce, K-Means, Fuzzy C Means.
5
Techniques Impact: After performing a survey with many
researches we have concluded Map Reduce or
Fuzzy C means are the best method for
processing large amount of data. It helps to select
proper parallel clustering algorithm.
 Finally, By performing more than 100+ literature
survey many researches in E-Commerce or M-
Commerce it has been reviewed many researchers
has developed optimization techniques to find the
best recommendations. Some have used parallel
clustering algorithms to process the data in efficient
way with reduced time. Some researched are
conducted to predict efficient outputs. Some
researchers have used genetic algorithms to fetch the
Prediction based Hybrid
relevant data. But no researches are conducted where
Optimization technique
6 combined all these methods are used with some new
for I-Commerce
researches to extract recommend the most relevant
output.
Impact: Above literature survey helps me to
select the appropriate method for selecting
appropriate parallel clustering method,
optimization algorithm and research area.
Hence, we have decided to conduct a research for
extracting the most relevant output using
Prediction based Hybrid Optimization technique
for I-Commerce
References and Books
A. Hammed and Ali Oudah, “Improved Methodology for Mobile Commerce Applications”,
International Journal of software Engineering and Its application, 2014.

A. Mitra, S. Ghosh, P. Basuchowdhuri, M. K. Shekhawat, and S. K. Saha, “Recommendation

system based on product purchase analysis,” Innovations in Systems and Software
Engineering, vol. 12, no. 3, pp. 177-192, 2016.

Akter, S., & Wamba, S. F. (2016). Big data analytics in E-commerce: a systematic review and
agenda for future research. Electronic Markets, 26(2), 173–194.

Andreou, Andreas & Chrysostomou, Chrysostomos & Leonidou, C & Mavromoustakos,

Stephanos & Pitsillides, Andreas & Samaras, George & Schizas, Christos. (2002). Mobile
commerce applications and services: a design and development approach.

Anon Sukstrienwong (2011) on “Searching Optimal Buyer Coalition Structure by Ant Colony
Optimization” in International Journal Of Mathematics And Computers In Simulation, Issue 4,
Volume 5, 2011, 352-360

Argade D. and Chavan H. (2015), “Improve Accuracy of Prediction of User’s Future M-

Commerce Behaviour”, Procedia Computer Science, Vol. 49, pp. 111 – 117.

Asghar Afshar Jahanshahi, Alirega Mirzaie, Amin Asadollahi, “Mobile Commerce Beyond
Electronic Commerce: Issues And Challenges”, Asian Journal of Business Management
Science(2015).

Avasthi S. and Dwivedi A. (2013), “Prediction of Mobile User Behavior Using Clustering”,
International Journal of Scientific and Research Publications, Vol. 3, Issue. 2, pp. 1-5.

Bahriye Akay, Dervis Karaboga “A modified Artificial Bee Colony algorithm for real-parameter
optimization” in Information Sciences—Informatics and Computer Science, Intelligent Systems,
Applications: An International Journal, Volume 192, June, 2012 Pages 120-142

Bucki, R., & Suchánek, P. (2012). The Method of Logistic Optimization in E-commerce. J. UCS,
18, 1238-1258.

C. Wang, Z. Deng, J. Lai and P. S. Yu, "Serendipitous Recommendation in E-Commerce Using

Innovator-Based Collaborative Filtering," in IEEE Transactions on Cybernetics, vol. 49, no. 7, pp.
2678-2692, July 2019.

Cavazos J., Moss E. B. and Boyle M. (2006), “Hybrid Optimizations: Which Optimization
Algorithm to Use?”, retrieved on 9th March 2017 and from
https://www.eecis.udel.edu/~cavazos/cc-2006.pdf

Chander (2009) “Optimization Techniques” by New Age Science (April 15, 2009)
Charoensiriwath, Supiya & Bentley, Peter. (2003). Particle Swarm Optimization Recommender
System. 124 - 131.

D. Sethi and A. Singhal, "Comparative analysis of a recommender system based on ant colony
optimization and artificial bee colony optimization algorithms," 2017 8th International Conference
on Computing, Communication and Networking Technologies (ICCCNT), Delhi, 2017, pp. 1-4.

Dawen Xia, Binfeng Wang, Yantao Li, Zhuobo Rong, and Zili Zhang, “An Efficient MapReduce-
Based Parallel Clustering Algorithm for Distributed Traffic Subarea Division,” Discrete Dynamics
in Nature and Society, vol. 2015, Article ID 793010, 18 pages, 2015.

Degree project in Computer Engineering, first cycle stockholm, Sweden 2017 by Utsav Kaushik ,
Antonio Grondowski (2017) on “Conversion rate optimization of E-commerce using web analytics
and human-computer interaction principles”

Ding, Yanping. (2014). Research on TSP Problem in E-Commerce Tourist Based on Ant Colony
Algorithm. International Journal of u- and e-Service, Science and Technology. 8. 277-288.
10.14257/ijunesst.2014.7.3.24.

Eastin, M. S., Brinson, N. H., Doorey, A., & Wilcox, G. (2016). Living in a big data world:
Predicting mobile commerce activity through privacy concerns. Computers in Human Behavior,
58, 214–220.

Evangeline S. J., Subramanian K. M. and Venkatachalam K. (2013), “Survey on Personal Mobile

Commerce Pattern Mining and Prediction”, International Journal of Advanced Research in
Computer Engineering & Technology (IJARCET), Vol. 2, Issue. 12, pp. 3163-3167.

Evangeline S. J., Subramanian K. M. and Venkatachalam K. (2014), “Efficiently Mining the

Frequent Patterns in Mobile Commerce Environment”, Vol. 5, No. 1, pp. 30-39.

Fahim Akhter, “Implement Fuzzy Logic to Optimize Electronic Business Success” International
Journal of Advanced Research in Artificial Intelligence(IJARAI), 5(3), 2016.

Feng, Zhitan. (2019). Constructing rural e-commerce logistics model based on ant colony
algorithm and artificial intelligence method. Soft Computing.

Georgiadis, C.K., & Manitsaris, A. (2005). Personalized Recommendations in Mobile Commerce.

Goswami, A., Zhai, C., & Mohapatra, P. (2018). Towards optimization of e-commerce search and
discovery. CEUR Workshop Proceedings, 2319.

Govind Kumar Sunda, Prof. Vaidhehi V (2017) on Search Engine optimization techniques for E-
Commerce in 2017, International Journal of Innovative Research in Advanced Engineering, AM
Publication 2017

Hans H Bauer, Stuart J Bauer, Tina Reichardt, Marcus M Newmann, “Driving Consumer
Acceptance of Mobile Marketing: A Theoretical Framework and Empirical Study”, Journal Of
Electronic Commerce Research, 2005.
Ho, D. T., & Garibaldi, J. M. (2013, July). An improved optimisation framework for fuzzy time-
series prediction. In Fuzzy Systems (FUZZ), 2013 IEEE International Conference on (pp. 1-8).
IEEE.

Huang, Xuan & Song, Zhijun. (2014). Clustering Analysis on E-commerce Transaction Based on
K-means Clustering. Journal of Networks. 9. 10.4304/jnw.9.2.443-450.

Humphreys (2016), “A Hybrid Optimization Technique Applied to the Intermediate-Target

Optimal Control Problem”, Global Journal of TechnolOptim, an open access journal, Vol. 7, Issue.
2, pp. 1-8.

International Journal of advanced Research in Computer and Communication Engineering ISO

3297:2007 CertifiedVol. 6, Issue 12, December 2017

Jhila Nasiri ,Farzin Modarres Khiyabani, Akiko Yoshise “A whale optimization algorithm (WOA)
approach for clustering” in journal of Cogent Mathematics & Statistics ,Volume 5, 2018 - Issue 1.

Jiang, L., Cheng, Y., Yang, L. et al. “A trust-based collaborative filtering algorithm for E-
commerce recommendation system” in J Ambient Intell Human Comput (2019).

Jing Zhang, Gongqing Wu, Xuegang Hu, Shiying Li and Shuilong Hao (2013) “A Parallel
Clustering Algorithm with MPI – MKmeans” in journal of computers, vol. 8, no. 1, january 2013

Kandasamy K. and Kumar C. S. (2015), “Modified PSO Based Optimal Time Interval
Identification for Predicting Mobile User Behaviour (MPSO-OTI2-PMB) in Location Based
Services”, Indian Journal of Science and Technology, Vol. 8 (s7), pp. 185–193.

Kim W. (2009), “Parallel Clustering Algorithms: Survey”, retrieved on 9th March 2017 and from
https://s3-us-west-2.amazonaws.com/mlsurveys/46.pdf

Kwon, O., Lee, N., & Shin, B. (2014). Data quality management, data usage experience and
acquisition intention of big data analytics. International Journal of Information Management,
34(3), 387–394.

Lavuri R. (2016) “E-commerce Progress in India and Consumer Perception Towards Online
Shopping”, Indian Journal of Applied Research, 6(8), pp 357–362

Lee S. and Ahn H. (2011), “The hybrid model of neural networks and genetic algorithms for the
design of controls for internet-based systems for business-to-consumer electronic commerce”,
Expert Systems with Applications, Vol. 38, pp. 4326–4338.

Li, Y. & Chung, S.M. “Parallel bisecting k-means with prediction clustering algorithm” J
Supercomput (2007) 39: 19.

Lucy Gitau, Dr David Nzuki, “Analysis of Determinants of M-commerce Adoption by Online

Consumers, International Journal of Business, Humanities and Technology”, 2014.
Ly-Fie-Sugianto, Zhigang Liao, Dewi Rooslini Tojib, “Unraveling The Inhibiting Factors Of M-
Commerce Adoption”, Monarsh University Research Publication, 2008

M. Nilashi, O. B. Ibrahim, N. Ithnin, and R. Zakaria, “A multi-criteria recommendation system

using dimensionality reduction and Neuro-Fuzzy techniques,” Soft Computing, vol. 19, no. 11,
pp. 3173-3207, 2015.

M. Seyedali, Seyed Mohammad Mirjalili, and Andrew Lewis. "Grey wolf optimizer." Advances
in engineering software vol. 69, pp. 46-61, 2014.

Maalini D. and Subashini S. (2014), “An Enhanced Prediction of Subsequent Mobile User
Behavior in Location Based Service Environments”, International Journal of Emerging Research
in Management &Technology, Vol. 3, Issue. 2, Pp. 55-60.

N. Cha, H. Cho, S. Lee and J. Hwang, "Effect of AI Recommendation System on the Consumer
Preference Structure in e-Commerce: Based on Two types of Preference," 2019 21st International
Conference on Advanced Communication Technology (ICACT), PyeongChang Kwangwoon_Do,
Korea (South), 2019, pp. 77-80.

N. Greenquist, D. Kilitcioglu and A. Bari, "GKB: A Predictive Analytics Framework to Generate

Online Product Recommendations," 2019 IEEE 4th International Conference on Big Data
Analytics (ICBDA), Suzhou, China, 2019, pp. 414-419.

N. S. Sani, and F. N. Tabriz, "A New Strategy in Trust-Based Recommender System using K-
Means Clustering," International Journal Of Advanced Computer Science And Applications, vol.
8, no. 9, pp. 152-156, 2017.

Nayak M. and Prakash K. B. (2013), “A Framework for Personal Mobile commerce Pattern
Mining and Prediction”, International Journal of Engineering Trends and Technology (IJETT),
Vol. 4, Issue. 9, pp. 3921-3925.

Niknam T., Nayeripour M. and Firouzi B. B. (2009), “Application of a New Hybrid optimization
Algorithm on Cluster Analysis”, International Journal of Electrical and Computer Engineering,
Vol. 4, No. 4, pp. 238-243.

Oreski S. (2014), “Hybrid Techniques of Combinatorial Optimization with Application to Retail

Credit Risk Assessment”, Artificial Intelligence and Applications, Vol. 1, No. 1, pp. 21-43.

P. Chu and S. Lee, "A novel recommender system for E-commerce," 2017 10th International
Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-
BMEI), Shanghai, 2017, pp. 1-5.

P. Pandey, S. Gudla and J. Bose, "Item recommendation in aggregated marketplace," 2019 16th
IEEE Annual Consumer Communications & Networking Conference (CCNC), Las Vegas, NV,
USA, 2019, pp. 1-3.
P. Papamichail, Georgios & Papamichail, Dimitris. (2007). The k-means range algorithm for
personalized data clustering in e-commerce. European Journal of Operational Research. 177.
1400-1408.

Panigrahi, A.K. and Joshi, V. (2016) “Application of E-commerce in Indian Textile Industry”.
International Journal of Textile Engineering and Processes, 2(2), pp.9- 12.

Panigrahi, Ashok & Joshi, Vijay. (2016). Application of E-Commerce in Indian Textile Industry.

Papamichail (2007), “The k-means range algorithm for personalized data clustering in e-
commerce”, European Journal of Operational Research, Vol. 177, pp. 1400–1408.

Perera Srinath (2013) “Hadoop MapReduce Cookbook” 2nd edition by PACKT publisher.

Pim Soonsawad (2013) Developing a New Model for Conversion Rate Optimization: A Case
Study, International Journal of Business and Management, Vol. 8, No. 10 (2013)

Prabir C. and Kumar M. (2017) “E-commerce: Promoting Entrepreneurship in India”, SCIREA,

1(2), pp. 43-54

Priyadharshini R. and Kalimuthu M. (2013), “An Efficient Framework for Predicting and
Recommending M-Commerce Patterns Based on Graph Diffusion Method“, International Journal
of Computer Applications Technology and Research, Vol. 2, Issue. 3, pp. 311 – 314.

Priyadharshini R., Geethanjali N. and Dhivya B. (2013), “Mining and Predicting Users M-
Commerce Patterns using Collaborative Filtering Algorithm”, International Journal of Engineering
Research and Applications (IJERA), Vol. 3, Issue. 3, pp. 238-242.

R. Dong, M. P. O’Mahony, M. Schaal, K. McCarthy, and B. Smyth, (2016) “Combining similarity

and sentiment in opinion mining for product recommendation,”Journal of Intelligent Information
Systems, Vol. 46, no. 2, pp. 285-312.

Rana, S., Jasola, S. and Kumar, R. (2011) “A review on particle swarm optimization algorithms
and their applications to data clustering” Artif Intell Rev 35: 211

Santucci V., Milani A. (2011) Particle Swarm Optimization in the EDAs Framework. In: Gaspar-
Cunha A., Takahashi R., Schaefer G., Costa L. (eds) Soft Computing in Industrial Applications.
Advances in Intelligent and Soft Computing, vol 96. Springer, Berlin, Heidelberg

Shao K., Cheng Y. (2010) E-Commerce Comparison-Shopping Model of Neural Network Based
on Ant Colony Optimization. In: Luo Q. (eds) Advances in Wireless Networks and Information
Systems. Lecture Notes in Electrical Engineering, vol 72. Springer, Berlin, Heidelberg

Suchanek, Petr & Bucki, Robert. (2011). Method of supply chain optimization in E-commerce.
Journal of Applied Economic Sciences.
Thomas, L., Manjappa, K., Annappa, B., and Reddy, G. R. M. (2011). Parallelized K-Means
clustering algorithm for self-aware mobile ad-hoc networks. Proceedings of the 2011 International
Conference on Communication, Computing & Security - ICCCS ’11.

Ulzhan Kakenova, Khaled Mohammad (2015) “Integrating Context Aware Analysis for M-
Commerce Application Prototype” , University of Bern, Switzerland Institute of Applied
Mathematics and Computer Science.

V.Dinesh , P.Deepika, S.Prabhu (2018) “Internet of Things (IoT) For E-Commerce - A Study”
International Journal of Innovative Research in Science, Engineering and Technology, Volume 7,
Special Issue 8, September.

Wijaya, h.c., and handhayani, t. (2017). Customer segmentation menggunakan fuzzy c-means
clustering pada e-commerce henz collection.

Witten, I. H., Frank, E., Hall, M. A., and Pal, C. J. (2016). Data Mining: Practical machine learning
tools and techniques. Morgan Kaufmann.

Wu, P.-J., & Lin, K.-C. (2018). Unstructured big data analytics for retrieving e-commerce logistics
knowledge. Telematics and Informatics, 35(1), 237–244.

X. Zhao, (2019) "A Study on E-commerce Recommender System Based on Big Data," IEEE 4th
International Conference on Cloud Computing and Big Data Analysis (ICCCBDA), Chengdu,
China, 2019, pp. 222-226.

Xiaobo Li, Zhixi Fang, (1989 )“Parallel clustering algorithms”, Parallel Computing, Volume 11,
Issue 3, Pages 275-290

Xiaoming Xu, (2014), IOT Technology Research in E-Commerce, Information Technology

Journal, 13: 2552-2559

Xiaowei Xu, Hans-Peter Kriegel, Jochen Jager (1999, “A Fast Parallel Clustering Algorithm for
Large Spatial Databases” in Data Mining and Knowledge Discovery, 3, 263–290

Xuan HUAN, Zhijun Song (2014), “Clustering Analysis on E-commerce Transaction Based on K-
means Clustering” in Journal Of Networks, Vol. 9, No. 2, February.

Y. Xie, Y. Ma, C. Ling and G. Wang (2012), "A novel parallel clustering algorithm PXM based
on FP-Tree," 2012 International Symposium on Instrumentation & Measurement, Sensor Network
and Automation (IMSNA), Sanya, pp. 475-481.

Y. Xu, Q. Yu, W. Lam, and T. Lin, (2017), “Exploiting interactions of review text, hidden user
communities and item groups, and time for collaborative filtering”, Knowledge and Information
Systems, Vol. 52, no. 1, pp. 221-254.

Yang, Xin-She & Deb, Suash and Fong, Simon. (2012). Accelerated Particle Swarm Optimization
and Support Vector Machine for Business Optimization and Applications. Communications in
Computer and Information Science.
Yu, H., & Zhang, X. (2017). Research on the Application of IoT in E-Commerce. 22017 IEEE
International Conference on Computational Science and Engineering (CSE) and IEEE
International Conference on Embedded and Ubiquitous Computing (EUC)
Websites
[1] https://blog.gluelabs.com/how-recommendation-systems-work-in-ecommerce-
6cc30b56b401 accessed on 9th April 2019
[2] https://lucidworks.com/post/clustering-classification-supervised-unsupervised-learning-
ecommerce/ accessed on 10th April 2019
[3] http://mdolab.engin.umich.edu/content/nature-inspired-optimization-algorithms accessed
on 9th June 2017
[4] https://en.wikipedia.org/wiki/Natural_evolution_strategy accessed on 25th July 2018
[5] https://en.wikipedia.org/wiki/Genetic_algorithm accessed on 25th July 2018
[6] https://en.wikipedia.org/wiki/Evolutionary_algorithm accessed on 25th July 2018