You are on page 1of 7

CHAPTER-1

Technical Report
on
Using IOT Technology to Improve Online Education
Through Data Mining

ABSTRACT: online learning and teaching. This paper provides a


proposal of a system where data collected from the
IoT devices will be used to help the educational
Online learning is rapidly expanding in both
institutions management make informed decisions.
the number of programs and users. Several data-mining algorithms will be applied within
Concerns are arising on how to improve the the central database systems to sort, arrange, merge
quality of materials and delivery methods. and analyze information to produce reports for
This research intends to explore the various levels of management. The real-time stream
of data and analysis of these transactions can be fed
potential use of Internet of Things (IoT) to
directly to a learning analytics system for data-driven
collect data, which will be analyzed to get decision making by the higher learning institutions.
information useful for decision making to The online access of the academic platforms are also
improve the online programmes in Higher affecting the administration of the universities and
Education Institutions (HEIs) colleges.

INTRODUCTION: OBJECTIVES:
Many people who are advancing their careers and The aim of this study is to propose a system
students are choosing online education to develop for collecting data from IoT devices and
their profession while still maintaining their job analyze it to help in improving the quality
status or seek training or prepare for certification
of information delivered by tutors and
exams. Due to the increase in number of users
registering for the online courses a close monitoring accessed by the learners. The research is
and evaluation by the institutions is becoming a intended to focus on the participants using
struggling issue[1]. The number of supervisors/ the online platform. The main objectives of
tutors doesnt cater with this increase, therefore the
system are:-
educational institutions are investing huge amount in
online teaching and learning in order to take the
(i.) To engage participant in online study and
advantage of the network as a platform. The Internet
offers high-bandwidth suitable for applications such provide them with an opportunity to
as animations, video conferencing, and simulations explore on various teaching and learning
to a worldwide audience of interconnected learners. methodology
The platform is providing an environment for sharing
of contents with each other and online course (ii.) Evaluate and measure the impact
modules[2]. IoT instruments are being used to collect
that online platform have on learners
data as the learners are interacting with the online
learning platforms such as Massive Open Online understanding of course being delivered.
Courses (MOOCS) and send it to a central database Encouraging the teachers and learners to
system for analysis. partake in the study process to improve the
The aim of this research is to explore whether delivery and access of the online platform.
Internet of Things (IoT) can be used to improve
RELATED WORK: signals, temperature, sweating, eye movement etc.
We expect a correlation between these types of data
The adaption of new technologies for learning has collected from the IoT devices and understanding of
thrived in the emerging of world of the Internet of the students. The data types can be changed at any
Everything (IoE)the networked connection of time if these indicators were not useful. Next, we
people, process, data, and thingswhich is collect information related to the understanding and
becoming the basis for the Internet of Learning comfort of the user using a questionnaire because
Things [4], [5]. Several researches have been carried this information is not measurable by the IoT
in the field of E-Learning on how to improve the devices. The questionnaire are designed to collect
learners performance academically and to sustain information in time slices not in general. This will
the institutions reputation on education. enable the users to respond during those intervals
In the UK, eight schools participated to find out how rather than waiting until the end of session. The
Internet of Things can enhance learning in science, users might not recall everything what happened
technology, and Geography in a scheme worth during the session if allowed to fill in the
$1.2M [6]. Numerous e-learning platforms are questionnaire at the end of lecture. For one use we
available today and most popular platforms are the use Open source Software (OSS) to find the relation
commercial systems Blackboard, Clix, and between the data collected from the IoT devices and
Desire2Learn, and the open-source platforms ILIAS, the questionnaire. Then build a model that will help
Moodle, OLAT, and Sakai[7]. With modern us repeat the experiment on large scale and collect
improvement of technology many learning massive data on the understanding of the users from
institutions are applying the Internet of Things (IoT) the IoT devices. We feed this data into a special
within their online learning platforms to collect, engine that can provide the worst scenario of
store and send the data to the central database understanding by comparing the time of the
system. The data collected is more complex and slides/lecture activities with the time of the sensor
challenging due to increase in the volume of data data and therefore can give advice.
collected[8], [9]. SYSTEM ARCHITECTURAL
IoT driven business transformation through seeking DESIGN:
opportunity in the context of product is used not just The proposed application will be developed using
in the product but also in achieving transformative object programming language Java and analysis
outcomes through fleet leverage not directly through engine (Hadoop). The architecture has four modules
advanced IoT functionality[10]. IoT helps industries as shown in fig 1 (i) Sensors devices and data
create value in the ecosystem through taking the concentration center (ii) Questionnaire (iii) Data
advantage of the available opportunities. It also analysis (iv) Data visualization. The application will
provides the industries with better understanding of apply different algorithms to give different reports
the IoT landscape through different ways so that for the management. Hadoop, is a popular open
value can be created and captured[11]. source implementation of the Googles MapReduce
ICT Initiatives and Implementation strategy of IoT has model, which was primarily developed by Yahoo.
gained increased performance as a new technology- Different industries such as Amazon, Facebook
driven research approach in e-Learning. The employ Hadoop to manage massive amount of data
information collected and analyzed allows on a daily basis from their websites. Apart from Web
instructor(s) to have deeper insights into learning. data-intensive applications, scientific data-intensive
applications (e.g., seismic simulations and natural
RESEARCH AND language processing) take maximum benefits from
EXPERIMENTAL METHODOLOGY: the Hadoop system[13], [14]. It consists of two major
This study will adopt an exploratory route and apply components Hadoop MapReduce engine and
quantitative and qualitative combination Hadoop Distributed File System (HDFS). HDFS divides
methodology approach. The datasets will be files
collected using the sensors devices and will be time- into blocks that are replicated among several
stamped to indicate the time they were collected for different computing nodes with no attention to
accuracy and completeness. The data collected will whether the blocks are divided evenly[15]. In
be analyzed using the R language to produce the educational data mining predictive modeling is
reports and simulate the algorithms. Some of the usually used in predicting performance. Our system
data types to collect includes: - Heart beat rate, brain
will apply predictive modeling that applies 4. Privacy and protection problem in data
classification, regression and categorization to management of IoT
produce the reports. There are several algorithms
under classification task that have been applied to 5. Data archive, index, scalability and access
predict students performance. Among the control for IoT data.
algorithms used are Decision tree, Artificial Neural
Networks, Naive Bayes, K-Nearest Neighbor and 6. Data warehouse and its query language for
Support Vector Machine. The Nave Bayes (NB) is the multidimensional analysis.
simplest and most commonly used algorithm for
classification. 7. Interoperability and semantic intelligibility
A. Sensor Data Collection and for heterogeneous data of IoT.
transmission:
The heterogeneous sensors will be used to collect
data, store and transmit to the centralized database. C. Cloud Processing, Analytics and
At this stage a special care should be taken into Visualization:
consideration in order to collect accurate data that Data analysis and visualization is fundamental to
includes its time stamp. In addition energy-efficiency, learning institutions to understand the hidden data
scalability and fault-tolerance of the sensors as this pattern in data that can lead to improved policies
will allow the management and distribution of the and recommendations for better planning,
resources across the whole framework structure. monitoring, evaluation, cost efficiency, and thus
For example adopting the data aggregation will improving quality of the learners. In this regard, we
reduce the amount of data transmission and thus will create an algorithm that will be able to do data
promoting the utilization of energy of sensor nodes. processing, analyze and visualize for the
B. Data Concentration Center: management. All the three levels will be interlinked
The heterogeneous sensors will capture and store a to allow communication and flow of data as shown
massive data before transmitting it. Therefore, it is in the fig. 1. The benefits that the users will gain
from using the visualization tools are improved
necessary to consider how to manage data collected
decision-making, improved
effectively and how to implement an algorithm that
collaboration/information sharing, Better ad-hoc
will be able to analyze and process data
data analysis and time saving.
conveniently. The data has the following
characteristics: (i) Data of smart objects tends to The result of the model of sensor data will inform
have its own implicit semantics (ii) Data of smart the understanding of the students. If the percentage
objects is likely to inaccurate, and usually is time- of the majority users indicate the low understanding
related or location-related (iii) In the environment of the contents, instructor is informed to change
of IoT, devices such as sensors and RFID will produce during the break sessions.
massive data streams.
The characteristics of IoT data put forward new CONCLUSION:
demands for data management and data mining
The objective of using the IoT is to provide a
that includes:-
framework to capture, store and transmit the data
1. Identification and addressing of smart to the centralized database. When the data is
objects as they will be large number of smart analyzed it will be able to inform the management
devices connected and needs identification and on areas that needs to be improved in order to
addressing effectively. This will enable to query improve learners quality of lifelong learning. With
or communicate with the smart objects. the advancement of technologies, higher learning
institutions requires to take the advantageous
2. Time-series level and event level data opportunities to be able confront many challenges,
aggregation such as delivery methods, quality of contents,
teachers learning leadership, pedagogical theory,
3. Data abstraction and compression for educational technology leadership, educational
filtering redundant data. structures and ideology.
and teaching. This paper provides a
proposal of a system where data
collected from the IoT devices will
be used to help the educational
institutions management make
informed decisions. Several data-
mining algorithms will be applied
within the central database systems
to sort, arrange, merge and analyze
information to produce reports for
various levels of management. Many
people who are advancing their
careers and students are choosing
online education to develop their
profession while still maintaining
their job status or seek training or
prepare for certification exams. Due
to the increase in number of users
registering for the online courses a
close monitoring and evaluation by
the institutions is becoming a
struggling issue[1]. The number of
supervisors/ tutors doesnt cater
with this increase, therefore the
CHAPTER-2 educational institutions are investing
huge amount in online teaching and
SEMINARNOTES learning in order to take the
Online learning is rapidly expanding advantage of the network as a
in both the number of programs and platformThe main objectives of
users. Concerns are arising on how to system are:
improve the quality of materials and i. To engage participant in online study
delivery methods. This research and provide them with an
intends to explore the potential use opportunity to explore on various
of Internet of Things (IoT) to collect teaching and learning methodology.
data, which will be analyzed to get
information useful for decision SYSTEMARCHITDESIGN:
making to improve the online The proposed application will be
programmes in Higher Education develop using object programming
Institutions (HEIs) language java and analysis engine
The aim of this research is to explore (HADOOP). The architecture has four
whether Internet of Things (IoT) can modules
be used to improve online learning i. Sensors devices and data
concentration center
ii. Questionnaire
iii. Data analysis
iv. Data visualization
v. A. Sensor Data Collection and
transmission:
vi. The heterogeneous sensors will be used to
collect data, store and transmit to the
centralized database.
vii. B. Data Concentration Center:
The heterogeneous sensors will capture and
store a massive data before transmitting it.
Therefore, it is necessary to consider how to
manage data collected effectively and how to
implement an algorithm that will be able to
analyze and process data conveniently. The
data has the following characteristics: (i)
Data of C. Cloud Processing, Analytics
and Visualization:
i. Data analysis and visualization is
fundamental to learning institutions to
understand the hidden data pattern in data
that can lead to improved policies and
recommendations for better planning,
monitoring, evaluation, cost efficiency, and
thus improving quality of the learners.
smart objects tends to have its own implicit
semantics (ii) Data of smart
ii. objects is likely to inaccurate, and usually is
time-related or location-related (iii) In the
environment of IoT, devices such as sensors
and RFID will produce massive data streams
ublic/b9/c6/ b9c609e0-deb9-479a-bc19-
94770297fd3a/14cs4580- iot_whitepaper-
charts-r2.pdf
[3] G. P. Anthony, The Evolution of Big Data
REFERENCES: and Learning Analytics in American Higher
Education, J. Asynchronous Learn.
[1] J. S. Maureen Bowers, Neal Freyman, Networks, vol. 16, no. 3, pp. 9 20, 2012.
Elizabeth McLellan, Brian Paxton, Online [Online].Available:
and Traditional Lectures: Evaluating Effects http://onlinelibrary.wiley.com/doi/10.1002/
of Social Presence and Learner Control, 978111860 0146.fmatter/pdf
Thesis submitted to the faculty of the [4] M. Selinger, A. Sepulceda, and J. Buchan,
Gemstone, University of Maryland, College Education and The Internet Of Everything,
Park, 2013. [Online]. Available: Cisco, p. 15, 2013. [Online]. Available:
http://drum.lib.umd.edu http://www.cisco.com/c/dam/
[2] Learning Cisco, The Internet of en_us/solutions/industrie
Everything : Fueling Educational Innovation s/docs/education/ education_internet.pdf
The Internet of Everything : Fueling [5] P. Mongkhonvanit, V. N. Dieu, and N. Van
Educational Innovation Workforce Der Linden, Graduate Student, Assumption
Evolution, pp. 1 11,2014. University of Thailand Applications of
[Online].Available Internet of Things in E-Learning, vol. 23, no.
http://globalstemalliance.org/media/filer_p 3, pp. 14, 2015. [Online]. Available:
http://www.ijcim.th.org/past_editions/2015 cultural things, Futur. Gener. Comput. Syst.,
V23N3/23n 3Page1.pdf vol. 66, pp. 187198, 2017.
[6]A. del Blanco, A. Serrano, M. Freire, I. [11] A. Van Schaack, Livescribe in K 12
Martinez- Ortiz, and B. Fernandez-Manjon, Education : Research Support, no. March,
E-Learning Standards and Learning 2009.
Analytics, 2013 Ieee Glob. Eng. Educ. Conf., [12] M. Piotrowski, Document-Oriented E-
pp. 12551261, 2013. Learning Components, no. April 2009, p.
[7] S. Cuomo, P. De Michele, A. Galletti, and 207, 2009. [Online]. Available:
F. Piccialli, A Cultural Heritage Case Study http://edoc.bibliothek.uni-
of Visitor Experiences Shared on a Social halle.de/servlets/DocumentServlet?id=720
Network, Proc. - 2015 10th Int. Conf. P2P, [13] T. White, Hadoop: The definitive guide,
Parallel, Grid, Cloud Internet Comput. vol. 54. 2012.
3PGCIC 2015, pp. 539544, 2016. [8] E. Van [14] P. Garg, Performance Enhancement of
Leemput, Internet of Things (IoT) Business Big Data Processing in Hadoop Map /
Opportunities Value Propositions for Reduce, no. 133050093, Masters thesis of
Customers, 2014. computer science & engineering Indian
[9] J. Holmstrm, IoT driven business institute of technology Bombay Mumbai pp.
transformation : Seeking opportunity and 1 28, 2014.
leveraging fleet advantages. [15] K. Shvachko, The Hadoop Distributed
[10] A. Chianese, F. Marulli, F. Piccialli, P. File System, IEEE 26th Symp. Mass Storage
Benedusi, and J. E. Jung, An associative Syst. Technol., pp. 110, 2010.
engines based approach supporting
collaborative analytics in the Internet of

You might also like