Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                

Gatwiri

Download as docx, pdf, or txt
Download as docx, pdf, or txt
You are on page 1of 16

DEVELOPING A COURSE RECOMMENDATION SYSTEM.

BUURI CHELSIE GATWIRI

CO27-01-1397/2019

PROJECT DOCUMENTATION SUBMITTED TO DEDAN KIMATHI UNIVERSITY


OF TECHNOLOGY IN PARTIAL FULFILLMENT FOR THE AWARD OF DEGREE
IN BACHELOR OF BUSINESS AND INFORMATION IN TECHNOLOGY.

MAY 2022.

1
DECLARATION

I do hereby declare without any reasonable doubt that the work in this document is my own
original work and it has not been previously presented to any University or considered for
certification.

NAME: BUURI CHELSIE GAWIRI

REGISTRATION NUMBER: C027-01-1397/2019

SIGN: ……………………….

DATE: ………………………..

This research proposal has been submitted with my approval as the university supervisor

DR JOHN WANDETO

SIGN: ………………………

DATE: ………………………

2
ABSTRACT
Universities and institutions are filled with at least 25% of students who are pursuing courses
they chose to do only because they had no choice; and not because of their skills nor interests
in life. Due to this, many end up not being street smart due to not applying academic
knowledge into the real world. The purpose of this study is to help students that have finished
their Kenya Certificate of Secondary Education (KCSE) examinations to make an informed
decision when choosing their career path. This course recommendation system is designed in
such a way that by use of questionnaires, the user is required to only tick answers that best
describe themselves, preferences, motivations and values in life. The main objective of this
study is to recommend respondents courses that best match with their interests, personality
and motivations based on the answers they provided. To accomplish this objective, the
questionnaires are set in a simple, precise format and multiple choices are also provided to
avoid unnecessary information from the user (closed questionnaire), unless their opinion on
the subject is required (open questionnaire). In addition, the system will have an extra feature
of recommending respondents their suitable institution to pursue their career based on the
course the system will have suggested or recommended for them.

3
CONTENTS
DECLARATION...................................................................................................................................2
ABSTRACT..........................................................................................................................................3
list of figures..........................................................................................................................................5
CHAPTER ONE: INTRODUCTION....................................................................................................6
1.1 Background of The Study......................................................................................................6
1.2 Purpose of The Study.............................................................................................................6
1.3 Statement of The Problem.....................................................................................................6
1.4 Objectives of The Study..............................................................................................................8
1.4.1 Main Objective.............................................................................................................8
1.4.2 Specific Objectives.......................................................................................................8
1.5 Research Scope......................................................................................................................8
1.6 Assumptions..........................................................................................................................8
1.7 Limitations of The Study.......................................................................................................8
1.8 Significance of The Study......................................................................................................8
1.9 Justification of The Study......................................................................................................9
CHAPTER THREE: METHODOLOGY............................................................................................10
3.1 Introduction...............................................................................................................................10

4
LIST OF FIGURES

Figure 1.3.1 1 Increase in ecommerce sales from 2014 to 2021............................................................7

5
CHAPTER ONE: INTRODUCTION
1.1 Background of The Study
Recommendation systems are rapidly becoming popular nowadays and sometimes people do
not even realize they exist. People seek recommendations on a daily basis for many purposes.
A recommendation system does not necessarily mean it should be an informatics system.
Through the understanding of a user’s preference, it is possible to create a suggestion for
them that will either satisfy them partially or fully (Estrela et al., 2017)
Universities offer a wide course selection to students studying in different fields. However,
this has generated various unintended problem areas. Students may choose a specific course
based on the reputation of the teacher, the course time, the course difficulty or ease, or other
factors not directly related to a specific field of study. Although they may achieve the
required number of credits to graduate, the breadth of knowledge they acquire may not equip
them well for the professional world they face after graduation.
For this work we propose a course recommendation system (CRS) that focuses on the relation
between course categories and students’ preferences when it comes to subject
recommendation. The proposed system has been developed to recommend suitable courses to
individual students, taking into consideration courses already taken, with the aim of
maximizing students’ professional foundation. We analysed a proportion of senior student
participants in order to evaluate the performance of the proposed system. Association rules
have been used to discover patterns in other databases. For example, university course
enrolment data were analysed to identify combinations of courses taken by groups of students
[3, 4]. After analysing a considerable volume of course attendance data, we applied an
association rule to understand the link between subjects chosen with those proposed by
faculty guidance advisors. We applied a minimum support and minimum confidence variable,
variously, and found an association rule. As the result of examining the validity of this rule,
we wish to apply to a revision enactment, or reference data, to the attending lecture tree.(Lee,
2011)
The recommender system will be used to analyse data automatically according to the user
preference and the most suitable one is presented in a plethora of different alternatives.
In India, a few online learning platforms or massive open online course (MOOC) like
national program or technology enhanced learning (NPTEL), active learning platforms for
young aspiring minds (SWAYAM) are developed by the government of India. These MOOC
can lead the learners to potential carrier failure without proper guidance and irrelevant course
chosen by the learners. A Course Recommendation System, CRS, may help the learner to
choose correct courses and the personalized environment will be able to engage the learner to
the framework.
This system will be an intelligent system that will recommend a personalized set of
information extracted from a dynamically generated huge volume of data.(Mondal et al.,
2020) Questionnaires provided will be based majorly on characteristics or basic personalities
of people who did or do similar courses.
Most of the universities and online platforms are widely offering internet-based learning
facility to the students. However in this situation, the most important concern is the
adaptability and understandability of the courses to the learners. Higher chances are that a
6
user may opt for an irrelevant course and lose the psych of study in the environment. A
learner-centric environment will help learners opt for the most suitable courses and by
personalization, will keep the learner engage to study and help to adopt skills by completing a
course faster.(Mondal et al., 2020)
For this work, three techniques were used to extract information and suggest recommended
courses: content based, collaborative filtering and hybrid.

1.2 Purpose of The Study


The purpose of this study is to try as much as possible to help students who get a hard time in
choosing and deciding which course they should do at higher learning institutions to avoid
making wrong decisions by getting a “helper” that will help them in making that decision.

1.3 Statement of The Problem


As the population of people in Kenya increases, the number of jobless people is not becoming any
better. Aside from other reasons that results to this, making the wrong decision when choosing a
career path is one of the leading factors. Individuals lack an informed decision when deciding
courses they’ll take and this affects them in future when they find out their energies are geared to
an entirely different field from what they learnt and specialized in in schools.

Most of them would term this as time wasted, which technically is not but on the other hand is;
since the course one decides to do at institutions should be a steering force to pursuing one’s career
in the future. Hence, it is important to have an idea of what one is capable of and use their strengths
to do something they would enjoy and not regret doing in future.

Due to this, there’s need to be a course recommendation system which would curb all this trouble
by simply answering a simple questionnaire that in corporation with psychology, masters the trend
of answers the respondents provide and match them with a course that best fits them.

7
1.4 Objectives of The Study.

1.4.1 Main Objective


The main objective of this study is to help form four leavers identify which courses would be
best fit for them based on their interests and preferences.

1.4.2 Specific Objectives


The specific objectives include:-

i. Exposure to career courses one may not have previously considered


ii. To help one in making a well-informed career decision
iii. To recommend users a suitable institution to pursue their careers based on the course
suggested for them
iv. To help one discover their interests, preferences, motivations and values in life.

1.5 Research Scope


The scope of this study will be all form four leavers in Nyeri town who have not yet joined
universities. The period of the research will take at least 4 months and will with no doubt benefit
everyone involved.

1.6 Assumptions
In this study it is assumed that:-

i. That all users are form four leavers and have not yet joined higher learning
institutions.
ii. That every respondent will have access to internet and also a smartphone.
iii. That every user taking the test is in the quest of knowing which course best match
with their traits.

1.7 Limitations of The Study


i. Dishonest and inaccurate responses from users who are not sure which traits they
possess.
ii. The system might not be able to cover all possible answers a user might think of
selecting.

1.8 Significance of The Study


The significance of this study to help form four leavers majorly, who with the right guidance
will also improve the society in one way or the other.

8
Learners

The personalized questionnaires set for learners to respond to will not only help them choose
a career to pursue but also help them discover their skills and motivations in life. The system
might also help learners consider courses they might not have previously thought of. In
addition, a suitable institution will also be recommended for them.

Academics

This sector will benefit hugely since as learners fill in the form, accurate courses are
recommended for them and this widens the institution field in various departments with at
least 96% of learners doing the course out of own will and passion.

Also, with majority of learners doing something they truly love, the academics sector will
save their capital and instead use it in more pressing issues.

Society

The CRS will benefit the society at large since with adequately learned personnel, there will
be an advancement in industries and organizations due to lower operating costs, greater
productivity, and also promote change and progress throughout the society and world at
large; which involves helping people with few or no resources.

1.9 Justification of The Study


The purpose of this study is to provide a platform where users can freely express themselves
by being honest to their interests and motivations in life.

9
CHAPTER TWO: LITERATURE REVIEW
2.1 Introduction
During the pandemic many businesses shifted to the online platform to do their business, in
this many customers were gained, to maintain their loyalty the business had to make sure
there customers are satisfied with the services they are offered. Many businesses adopted the
bot technology at this time.
In this section we will take a look at some of the few businesses that have automated their
services by using the bot technology. We will compare and contrast with their already
developed bot with the e-sales bot.
There is need for more learning of machine learning as the world is getting industrialized by
technology. Technology is advancing everyday hence a lookout so as not to be passed by the
current trending technologies.
2.2 Case Studies
This section includes a detailed study of similar or almost similar research studies done by
other people in an aim to reach the same objectives. It records all the research gaps in the
efforts of other researchers.

2.2.1 Chatbot by EBay


A popular and advanced example is the eBay chatbot. EBay’s chatbot, which was built for
the Google Assistant, may be utilized with Google Home or on the phone. Customer
questions about products will be answered by the bot, which will also drive the sales process.
It can, for example, respond to user queries about the greatest new products or the lowest
price options across all of eBay's global sites. The interface, which can be accessed by asking
"OK Google, let me talk with eBay," is designed to assist consumers in navigating a variety
of product categories. (eBayinc, 2016). Despite the eBay shopbot being integrated in
Facebook Messenger it is yet to be integrated to WhatsApp messenger where there is a
variety of users. All other functionalities that the shopbot has, E-Sales can perform the same.

Figure 2.2.1 1 An eBay shopbot.

10
2.2.2 Chatbot by LeadMD.
LeadMD is a revenue performance consulting firm that offers a variety of services, including
marketing automation, sales enablement, and CRM deployment. Their bot is quite helpful in
assisting consumers in quickly obtaining the information they require. LeadMD's chatbot
marketing techniques are enhanced by the LeadMD bot, which prioritizes the consumers'
time. The bot also provides a copy of the conversation, which is a one-of-a-kind feature
among others. Users can exchange their email addresses with the bot in order to help the
company's lead generating efforts. This functionality can also aid in the more efficient
tracking of a customer's journey. (OutGrow, 2022). The LeadMD bot offers sales and it is not
yet to be integrated in any social media platform and its main focus is in revenue consultancy.
The E-Sales bot will be integrated in WhatsApp where it will be accessible to all people and
it will be involved more in e-commerce than the revenue field.

Figure 1: An example of a LeadMD chatbot

2.2.3 Chatbot by Safaricom Zuri


Safaricom, a Kenyan mobile operator, has added an artificial intelligence chatbot assistance
to WhatsApp’s, Telegram’s and Messenger’s messaging service. Zuri allows Safaricom
mobile data subscribers to control subscription services and marketing messages, among
other things. PUK retrieval, M-Pesa reversals, buying bundles, accessing use statements,
airtime top-up, and monitoring M-Pesa and airtime balances are among the other issues Zuri
can handle. Customers may now access and examine their Safaricom Home Fibre Account
data through the chatbot's Safaricom Home area (Safaricom, 2018). Safaricom has turned to
chatbots to help with simple activities that don't require human interaction in an effort to
improve customer service delivery. The entire discussion is error-free because it is simply a
matter of following instructions; yet, the question arises as to why someone would want to
utilize a chatbot over other methods of obtaining the same services; the answer is simple: it is
easier. (Techguy, 2018). Safaricom Zuri is a Kenyan made bot, it deals with sales of data,
checking Mpesa transactions among other things. The E-Sales bot will be more involved in
selling computer items, cars among other things and answering queries by the customer.
Hence it will be more similar to Zuri only it will be integrated on WhatsApp only.

11
Figure 2: Zuri on Telegram and Messenger.

Figure 3: Zuri in WhatsApp.

2.2.4 Chatbot in Kilimall


Kilimall is an online platform where there is selling and buying of items. Kilimall has created
a bot to attend to customers when they inquire about transactions or other questions related to
the products they sell. Kilimall can be contacted via live chat on their website, or by sending
an email to cc@kilimall.com.We encourage contacting us via live chat, where you may share
your order and other details for faster and more efficient service. (Kilimall, 2019). Comparing

12
the kilimall live chat to E-Sales there is a big difference in it that, the bot will be trained with
the questions to answer and also it will be integrated in WhatsApp and no end –human
interaction will be required. When the customer care team goes home, no one is left to tend
for the customers’ needs.

Figure 4: Kilimall Messaging Platform.

2.3 Summary
From the cases above it is evident that there is advancement of technology. It can be observed
that most brands have adopted the technology of chatbot integration in their systems while
others are yet to. The use of automated chatbots have made work easier for all people that is
both customers and the sellers. Some may have automated their chatbots but they are yet to
integrate it into WhatsApp which is the commonly used messaging application. Some
systems are yet to be updated with the current trends in technology.
2.4 Research Gap

ALL THESE SYSTEMS ARE TO ENSURE EFFICIENCY ON THE SIDE OF


CUSTOMER SUPPORT TEAM AND ALSO TO REDUCE THE WORKLOAD FOR
THEM AND TO ENSURE THE CUSTOMER IS SERVED AT ALL TIMES WHEN
THE NEED ARISES.

13
CHAPTER THREE: METHODOLOGY
3.1 Introduction
Research methodology is a systematic way to solve a problem. It is a science of studying how
research is to be carried out. Essentially, it is the procedures by which researchers go about their
work of collecting, describing, explaining and predicting phenomena (Ketchen & Bergh, 2006). The
research is aimed at solving the inadequate exposure to the online CRS technology. Data that is both
qualitative and quantitive shall be collected from the subject of the population.

3.2 Research design


The nature of the research design was an applied approach which aims at finding a solution
for an immediate problem that is facing the society or a school institution. Various techniques
and methodologies were implored in an effort to achieve the anticipated objectives. Mainly,
the methods chosen particularly revolved around using existing knowledge so as to provide
better solutions to problem in learning, by creating a new system, which is the discipline in
this research.
The type of research applied in this study mainly involved descriptive, applied and qualitative
research types. In descriptive research type, surveys and techniques to find out facts were
used. There was no control over the variables present. Only reporting what exists. From this
method I was able to collect data about various properties, interests and motivations students
who pursue specific courses possess so as to formulate the questionnaires to be provided for
the system.
Applied research was also used where there was need to find a solution to a better schooling
experience and evade the probability of having to change a course and or having regrets in
the future. CRS had to be applied to help learners in making a decision of which course to
choose.
Lastly, qualitative research was applied when trying to find an effective way of matching
users with courses that accurately describe their personality and interests in life. Different
approaches were tried in an effort to find an improved and efficient way of gathering this
information.

3.3 Data Collection Methods


An important discipline in research studies is collection of data from the target populations
selected by a researcher. From the data that is collected inference and relationships can be
made from this that help any researcher to know the points to improve. Here are the methods
I used in this research study;
3.3.1 Questionnaires
A blend between closed and open questionnaires will be used to collect data from the selected
subject of study. The questionnaires will be deployed through google forms which will be
sent through the internet. It will help in the research of finding out traits possessed by
different people doing different courses. This is a chosen method because it’s a cheap,
effective and quick method of obtaining information from large population of respondents,
14
especially using google forms. Also, while using this method, I won’t be required to be
present where the respondent will be answering to the questions, hence their responses will
not be tampered with in an form or manner. The chances for bias in the data collected is less
while using questionnaire. Also, for privacy reasons, for those students who may not want to
reveal their identity, a questionnaire proved to be worthwhile. (Muhammad & Kabir, 2016)

3.3.2 Interviews
Interviews will be preferred where a particular number of students in the related field will
give their insight concerning the topic of study. Structured interviews will be used in this
case, where the students will be given a particular pre-defined number of questions. A
structured interview so that the respondent, in this case who is a form four leaver, will not
deviate from the topic chosen or else probe beyond the answers received. By standardizing
and pre-defining the questions, I will be able to make equitable comparisons between what
the respondents have given as answers.(INTERVIEW HANDBOOK Career Development Services,
n.d.)

3.4 Software Development Method


Agile development method will be used in this instance. Agile method proposes incremental
and iterative approach to software design. This results in small incremental releases with each
release building on previous functionality. This method is less rigid to change thus
incorporates change as development progress. The division of the entire project into smaller
parts helps to minimize the project risk and to reduce the overall project delivery time
requirements.(Dingsoyr et al., 2010)

Figure 5: Follow of Agile Method.

15
3.5 Sampling and Sampling Methods
Sampling refers to the selection of a number of study units from a defined study population.
It will determine what group of study will be selected to represent the whole field of study in
this research.
Non-probability Sampling will be used in this case as I have no control over who will be part
of the selected sample of study. Convenience sampling will be used as it enables one to draw
a sample from the population close at hand for pilot testing.
3.6 Software Used in Study.
Front End Development Python

HTML, CSS, JS

Backend Development- PHP

Course Recommendation System HTML, PHP, Python

Software Requirements Visual Studio Code

Xampp Server

3.7 Justification for Methodology


The main reasons why the researcher adopted this method is because: -
i. It continuously gives attention to technical excellence and good design.
ii. Regular adaptation to changing circumstances.
iii. People and interactions are emphasized rather than process and tools. Customers,
developers and testers constantly interact with each other.
iv. Even late changes in requirements are welcomed.

16

You might also like