nurul syafiqah izzati binti abdul hadi bachelor of...
TRANSCRIPT
DIABETES PREDICTION SYSTEM
NURUL SYAFIQAH IZZATI BINTI ABDUL HADI
BACHELOR OF COMPUTER SCIENCE
(SOFTWARE DEVELOPMENT)
UNIVERSITI SULTAN ZAINAL ABIDIN
2017
DIABETES PREDICTION SYSTEM
NURUL SYAFIQAH IZZATI BINTI ABDUL HADI
Bachelor of Computer Science (Software Development)
Faculty of Informatics and Computing
Universiti Sultan Zainal Abidin, Terengganu, Malaysia
MAY 2017
i
DECLARATION
I hereby declare that this report is based on my original work except for quotations
and citations, which have been duly acknowledged. I also declare that it has not been
previously or concurrently submitted for any other degree at Universiti Sultan Zainal
Abidin or other institutions.
________________________________
Name : Nurul Syafiqah Izzati Binti Abdul Hadi
Date : ..................................................
ii
CONFIRMATION
This is to confirm that:
The research conducted and the writing of this report was under my supervison.
________________________________
Name : Nor Surayati Binti Mohamad Usop
Date : ..................................................
iii
DEDICATION
First at all, be grateful to Allah SWT because got chance to finish the my final year
project, DIABETES PREDICTION SYSTEM. Thanks also to my supervisor,
MADAM NOR SURAYATI BINTI MOHAMAD USOP because willing to teach and
motivate me in order to finish this final project.This work is dedicated to my parent,
ABDUL HADI BIN MOHAMED and SABARIAH BT MD YUSUF, without whose
caring support it would not have been possible. Not forget also, my friends and my
classmate, thanks to them because together help me complete this project.
I am really appreciate their action to me.
iv
ABSTRACT
Diabetes Prediction System is the web based system to be used in medical field. This
idea is inspired because there are lack of awareness about diabetes disease among
people at this world.Diabetes can cause many worse diseases such as heart failure,
nerve damage, eyes problem and another organ failure. Next, the early diagnosis can
prevent the disease become more worse. This system are build to do early diagnosis.
In this system, doctors as user can predict their patients condition whether they will
having the diabetes disease or not based on their records. Next, the another user can
also access this system to get early diagnosis. Then, to if they want get more accurate
result, they can refer to specialists. So, this system main moduls are consist of user
and administrator. The user will provide the details about their health condition and
personal information to get the results.Then, the questionnaires is purposed and
answerscan be collected .Next, the system will provide early diagnosis result after do
the calculation and generate result based on the input. The system will use rule based
algorithms. Rule based algorithm can be used for powering prediction the disease and
are used to implement “IF THEN” in this system. PHP coding will develop into web
where there is system will predict the diabetes. Also,the tips and information about
Diabetes sections also available to be viewed by users. Through that, user able gain
knowledge about Diabetes.
v
ABSTRAK
Sistem Ramalan Diabetes adalah sistem berasaskan web yang akan digunakan dalam
bidang perubatan. Idea ini diilhamkan kerana terdapat kekurangan kesedaran
mengenai penyakit diabetes di kalangan orang-orang di world. Diabetes ini boleh
menyebabkan pelbagai penyakit kritikal seperti kegagalan jantung, kerosakan saraf,
masalah mata dan kegagalan organ lain. Seterusnya, diagnosis awal boleh mencegah
penyakit menjadi lebih teruk. Sistem ini dibina untuk melakukan diagnosis awal.
Dalam sistem ini, doktor sebagai pengguna boleh meramalkan pesakit mereka sama
ada mereka akan mempunyai penyakit kencing manis atau tidak berdasarkan rekod
mereka. Seterusnya, pengguna lain juga boleh mengakses sistem ini untuk
mendapatkan diagnosis awal. Kemudian, jika mereka mahu mendapatkan keputusan
yang lebih tepat, mereka boleh merujuk kepada pakar. Jadi, sistem ini moduls utama
adalah terdiri daripada pengguna dan pentadbir. Pengguna akan memberikan butiran
mengenai keadaan kesihatan mereka dan maklumat peribadi untuk mendapatkan
keputusan. Kemudian, soal selidik yang telah dilakukan dan jawapan dikumpulkan.
Seterusnya, sistem akan memberikan hasil diagnosis awal selepas melakukan
pengiraan dan menjana hasil berdasarkan input. Sistem ini akan menggunakan
algoritma berasaskan peraturan. algoritma berasaskan peraturan boleh digunakan
vi
untuk menjanakan ramalan penyakit ini dan digunakan untuk melaksanakan "IF
THEN" dalam sistem ini. PHP coding akan berkembang menjadi web di mana
terdapat sistem akan meramalkan diabetes. Juga,bahagian tips dan maklumat tentang
Diabetes juga boleh didapati dan dilihat oleh pengguna-pengguna. Melalui itu,
pengguna dapat memperolehi pengetahuan mengenai Diabetes.
vii
CONTENTS
PAGE
DECLARATION i
CONFIRMATION ii
DEDICATION iii
ABSTRACT iv
ABSTRAK v-vi
CONTENTS vii
LIST OF TABLES vii
LIST OF FIGURES xvi
LIST OF ABBREVIATIONS xv
CHAPTER I INTRODUCTION
1.1 Introduction( Project Background) 1-2
1.2 Problem statement 3
1.3 Objectives 3
1.4
1.5
Scopes
Expected Outcome
3-4
5
1.6 Limitation of Work 5
CHAPTER II LITERATURE REVIEW
2.1 Introduction 6
2.2 Research About Diabetes 6-7
2.3 Analyse Toward Existing System and Related with
others Method
8-9
2.4 Research About Rule Based 9-10
2.5 Conclusion 10
CHAPTER III
METHODOLOGY
3.1 Introduction 11
3.2 Justification Selection 11-12
3.3 Methodology 12-13
3.4 System Requirement 14
3.4.1 Software Requirement 14
viii
3.4.2 Hardware Requirement 15
3.5 Introduction Of System Modelling 16
3.6 Framework 17-18
Process Model 19-28
3.7 Context Diagram
3.8 Data Flow Diagram
3.9 Data Flow Diagram Level 1
3.9.1 Manage User
3.9.2 Manage Prediction
3.9.3 Manage Result
3.9.4 Manage Respond
3.9.5 Manage Admin
3.9.6 Manage Segment Info
19-20
20-,22
23
23
24
25
26
27
28
Data Model
3.10 Entity Relationship Diagram 29-30
3.11 Database Modelling 30
3.11.1 Table Admin 31
3.11.2 Table Respond 31
3.11.3 Table Info 32
3.11.4 Table Questionnaire 32
3.11.5 Table Result 33
3.11.6 Table User 33
3.12 Conclusion 34
REFERENCES 35-36
ix
LIST OF TABLES
TABLE TITLE PAGE
3.4.1 List Of Software 14
3.4.2 List of Hardware 15
x
LIST OF FIGURES
FIGURE TITLE PAGE
3.2 Spiral Model 12
3.6 Framework for Diabetes Prediction System 18
3.7 Context Diagram 19
3.8 Data Flow Diagram Level 0 for Diabetes Prediction
System
21
3.9.1 Data Flow Diagram Level 1 for Manage User 23
3.9.2 Data Flow Diagram Level 1 for Manage Questionnaire 24
3.9.3 Data Flow Diagram Level 1 for Manage Result 25
3.9.4 Data Flow Diagram Level 1 for Manage Feedback 26
3.9.5 Data Flow Diagram Level 1 for Manage Admin 27
3.9.6 Data Flow Diagram Level 1 for Manage Segment Info 28
3.10 Entity Relationship Diagram of Diabetes Prediction
System
29
3.11 Table in Database Diabetes Prediction System 30
3.11.1 Table Admin 31
3.11.2 Table Respond 31
3.11.3 Table Info 32
3.11.4 Table Questionnaire 32
3.11.5 Table Result 33
3.11.6 Table User 33
xi
LIST OF ABBREVIATIONS / TERMS / SYMBOLS
CD Context Diagram
DFD Data Flow Diagram
ERD Entity Relationship Diagram
xii
LIST OF APPENDICES
APPENDIX TITLE PAGE
A Gantt Chart FYP1 37
1
CHAPTER I
INTRODUCTION
1.1 Background
Most people already heard about Diabetes disease. However, still many people
looked down and take it easy about this disease. They assume that Diabetes just a
simple one disease and can be cured easily. But they totally wrong. Diabetes can be
cause another chronic disease such as heart failure, nerve damage, eye problem and
another organ failures. There are categorization of the diabetes such as Type 1
diabetes, Type 2 diabetes, diabetes that caused by another disease and Gestational
diabetes mellitus (GDM) [4].
At Malaysia, clinics under Kementerian Kesihatan Malaysia(KKM) was
established to trace improvement in treating diabetes at health clinic (KK) under
KKM. There are about 657,839 registered patient where 653,326 are positively have
Type 2 Diabetes Mellitus from 2009 until 2012 and females patients are higher than
males patients and the races which have many patients are from Malays [3].
However, the reason that leading the diabetes still become questions although
the potential cause such as obesity and unhealthy lifestyle can tend become the factor
[2]. Besides obesity and unhealthy lifestyle, the another reason such as family history,
smoking, suger intake daily, and so on. So, we advised to be careful in practise our
2
lifestyle. It maybe become health or unhealthy lifestyle. The choice in our hands. To
the obesity patients, are advised to change diet style and take nutrition food in right
portion and always do exercise. The diabetes able make us become disable people
where there is organ are imputated because of diabetes complication. So, diabetes
patient should give attention to their leg. Early preventation more better than cured
disease.
There are about fourty eight percent of patient above 30 years old did not
realise that they are diabetes patient [5]. Then, the early prediction of diabetes should
be done by everyone before its late. Early prediction can save many lives. So, we
purpose the Diabetes prediction system based rule based and tree decision algorithm.
In this project, we propose a rule based algorithm as decision support system(DSS)
for diabetes prediction system.
Next, this project will arranged as starting by Chapter 1 where it will introduce
the project, next, followed by chapter 2. In this chapter, there is literature review are
available to explain about another research. Then, Chapter 3, the project methodology
and design and modelling. Lastly, the chapter 4 will identify the results of the
purposed project.
As conclusion, this research is to establish the system that able to identify whether
the person are positively or negatively have diabetes disease.
3
1.2 Problem Statement
The problem about this project is it is not easy to do diagnosis whether it is
positive or negative having diabetes. It is because of many reason. Different people
maybe have different signs. So it is not easily to assume that they have it or not. The
sign of the diabetes is always thirsty, always hungry, weight become decrease, feel
weak, have problem of sight, headacnes, always do urination and so on. However, the
real diagnosis are still needed to assign the real result.
1.3 OBJECTIVES
a) To measure the probability of a user for getting diabetes.
b) To implement rule based algorithm as prediction technique into a system
c) To develop the system that function with the real problem.
1.4 SCOPE
This system will focus on the potential user, admin and system.
i. Potential user
The user that have signs or not can get early diagnosis about
diabetes and takes early preventation.
User need sign up and then sign in. They need to fill out the
personal information. Next, do prediction by using the system.
User also able to view preventation and information about
Diabetes section in this system.
4
ii. Admin
The person who will coordinate this system and update the
system based on situation.
People who responsible to update information section in the
system.
iii. System
Login
There is login and registeration to enter this system
based on type of user.
Questionaire module
There is questionaire that need be answered and
evaluation by potential user and from that, the result can
be find out.
Domain System(Diabetes)
The result can be find out after analyzing through rule
based and tree decision technique.
Opinion or its rate
User can give opinion and suggestion and give their rate
about the system. Example, the early prediction really
can be trusted or not.
Information Section
User can find out more information of Diabetes
Through the information, users who potentially have
Diabetes can take early preventation.
5
1.5 EXPECTED OUTCOME
The system is expected to give accurate prediction based on the sign are gived
by user. User can answer the questionaire based on the real signs they had. Then
system will print out result of the prediction. User also able give suggestion and rate
the system. This system also expected to give the accurate info which is based on the
profesional observation.
1.6 LIMITATION OF WORK
This system is only give the early prediction based on the signs that user had.
The result of prediction may not accurate like the diagnosis from doctor. User need to
seek consultant with doctor if want the real one diagnosis. This system just able to
alert user to take fast action about the diabetes. If the user are predicted have positive
diabetes, the system will give suggestion.
6
CHAPTER 2
LITERATURE REVIEW
2.1 INTRODUCTION
In this chapter, the research about the existing system and proposed system
will be disscussed. There is weakness will be found out after analyzing the previous
existing system. Next, it will explained how the Rule Based will be implement in the
proposed system. Besides that, rule based algorithm are widely used in medical field
and another fields. Research paper related about rule based in anaother disease cases
also analyzed.
2.2 RESEARCH ABOUT DIABETES
Diabetes are known the one of the top disease in this world. Diabetes are not
easily to be cured and need to depend on the medicine. If we know earlier that we had
diabetes, we can control its impacts become more worse. The people only know that
they have diabetes after the effects already become worse. So, the early prediction are
required to aware all the people. There are three type of Diabetes that has been
identified such as Type 1, Type 2 and Gestational Diabetes. All this type diabetes have
its differences and characteristics. Also, the majority patient who has Diabetes is
female than male [1].
7
Type 1, usually people who suffer this type diabetes, she/he cannot produce
insulin anymore in their body because the pancreas totally damaged. Futhermore, the
average people have this type of diabetes is below 20 years old [1]. Next,the patient
with this type have weight loss. However, this disease is not easily to classified
whether the patient can have this type diabetes or can become into Diabetes Type 2
[4].
As well as Type 1, Type2 Diabetes patient also have problem in producing
insulin for their body, where their pancreas produce insulin ,however it still no enough
because their body resistant toward insulin. The majority of the Diabeties patients had
this type diabetes [3]. For normal person, the sugar level cannot low or more from the
normal level which is from 4.4 until 6.1 mmol/L [1].
Gestational Diabetes, commonly the person who have this type diabetes are
consist of pregnant women. During pregnancy moment, the pregnant women are
advised to do a few test to check they have this kind of diabetes or not. If the person
have this diabetes, the production of insulin cannot be produce as usual as before
pregnant. The risk for the baby to suffer from diabetes also higher. For information,
usually the high weight baby maybe delivered by the Gestational Diabetes mother.
Next, for the next pregnancy, the patient have high risk to get the same problem. The
bad effect to pregnant women who have this diabetes is bleeding during birth or
miscarriage may occur [1].
8
2.3 ANALYSE TOWARD EXISTING SYSTEM AND RELATED WITH
OTHERS METHOD
There are many diabetes prediction system are available right now. All these
systems, they use many technique method which is involving data mining. Medical
prediction is the result based on classification method data mining. The classification
process able decreasing faults because of the duration of prediction [6]. About above
70 percent shown that the classification are working precisely.
The example of the classification method are Naïve Bayes, Tree Decision,
Fuzzy Logic, Neural Network and Fuzzy K-nearest Neighbour. The top one is Naïve
Bayes. However, it is complicated to be implemented than rule based algorithm.
Fuzzy K-nearest Neighbour is combination of Fuzzy and K-nearest Neighbour.
So, not wonder Fuzzy K-nearest Neighbour more powerful in doing precise prediction
than K-nearest Neighbour. The problem with K-nearest Neighbour is not able evaluate
the strongest of partnership in the class. So the research use Fuzzy K-nearest
Neighbour to solve the problem. However, this method not able generate the huge
amount of accurations [6].
As conclusion, the performance of algorithm is different based on machine
learning. Example, in TANAGRA, Naïve Bayes is the best with 100% accurate [8]. In
this project, we use WEKA as machine learning tool. So, with WEKA, rule based able
to be apply by us.
For this project, we use rule based after do some researches. We choose rule
based because of few factors. The factor are easily to understand and not complicated
9
like another algorithm. Next, it is better than Tree Decision algorithm. Lastly, rule
based already famous with its implementation in medical field.
2.4 RESEARCH ABOUT RULE BASED
Rule based system is the results of many rules. The popular method is IF as the
cause (antencendent) THEN as the effect (consequent). It also has been used in
artificial intelligence system application. “IF” as input while “THEN” as output. There
are many input ir condition, so “AND” and “OR” maybe added in the statement of “IF
THEN”. Rule based also can do anything work that related with classification,
regression and association. These rule can be simplify by using prunning method.
Single rule need to be prunned after it fully finished [10]. The advantage of using rule
based is it able to handle computational complexity in rule based model [10].
Example of rule based,
If in the exam, the student get 90% above, they will get A and if lower than
that and above 80%, they will get B.
IF 90 <= grade,THEN class=A.
IF 80<= grade AND grade<90 , THEN class=B.
Here the another rule,
R1: IF age=youth AND student=yes THEN buys computer=yes.
R2: (age=youth) ^ (student = yes))(buys computer = yes) [7].
10
We use ^ as AND while v as OR. So we can assume that the rule based is simple one
than another method.
2.5 CONCLUSION
In this chapter, a few collection literature review has been done. This literature
review helps us to understanding our technique that will used us to gain knowledge
about the another technique that has been used in the previous research. Lastly, we
gain knowledge about the diabetes disease and get awareness about this disease.
11
CHAPTER 3
METHODOLOGY
3.1 INTRODUCTION
In this chapter, it will focused on the methodology that being applied in the
software development. The methodology of software development is the method in
managing project development. There are many model of the methodology are
available such as Waterfall model,V model, Incremental model, RAD model, Agile
model ,Iterative model and Spiral model . However, it still need to be consider by
developer to decide which is will be used in the project. The methodology model is
useful to manage the project efficiently and able to help developer from getting any
problem during time of development. Also, it help to achieve the objective and scope
of the projects. In order to build the project, it need to understand the stakeholder
requirements.
3.2 JUSTIFICATION SELECTION
For this project, we purpose Spiral Model as the model of the methodology
,that has been widely applied in the other project. It is because of few reason. There
are many advantage of using spiral model, any idea can be added at later phase, the
budget of the system can easy to be predict and user can give their opinion anytime
[9]. This spiral model very suitable if there is any changes at another moment.
12
There are four phases that involved in the spiral model that including planning,
risk analysis, engineering and evaluation. For each phase, there are activities are
involved. In 3.3 section, there is explaination of the activity of each phase. Figure 3.2
below shown that the planning phase as the start and evaluation as last phase.
Figure 3.2 Spiral Model
3.3 METHODOLOGY
In the Diabetes Prediction System, Spiral model has been chosen as the
methodology .There are four phases that involve in the spiral model:
1) Planning phase
Phase where the requirement are collected and risk is assessed. This
phase where the title of the project has been discussed with project
13
supervisor. From that discussion,Diabetes Prediction System has been
proposed. The requirement and risk was assessed after doing study on
existing system and do literature review anout another existing
research.
2) Risk analysis Phase
Phase where the risk and alternative solution are identified. A
prototype are created at the end this phase. If there is any risk during
this phase, there will be suggestion about alternate solution.
3) Engineering phase
At this phase, a software are created and testing are done at the end this
phase.
4) Evaluation phase
At this phase, the user do evaluation toward the software. It will done
after the system are presented and the user do test whether the system
meet with their expectation and requirement or not. If there is any
error, user can tell the problem about system.
14
3.4 SYSTEM REQUIREMENT
Based on techopedia.com, the implementation that the system needed to make sure the
hardware or software can be run smoothly. If not success in fulling the requirement,
the failure of performance and installation may occur.
3.4.1 Software Requirement
The software requirement are needed to build system are:
Table 3.4.1: List of Software
SOFTWARE DESCRIPTION
XAMP Server MySQL Using this software to create database and
manipulate database and connect database
with PHP.
Edraw Max Create and design Data Flow Diagram
and Context Diagram
Dropbox Save and update the document for this
system and also as the backup file.
Google Chrome Medium to find reference to do system
and as medium to system be display and
run.
Notepad++ As medium to write PHP coding to build
system.
15
3.4.2 Hardware Requirement
The hardware requirement to build the system are:
Table 3.4.2 List of Hardware
HARDWARE DESCRIPTION
1) Laptop ASUS A550C
Have Intel i5 processor
4GB RAM
Window 8 operating system
64 bit Operating system type
16
3.5 INTRODUCTION OF SYSTEM MODELLING
By Kast and Rosenzweig, system is organized and complicated one. So,
system modelling able to assist analyst be capable in understanding functionality and
models of their system to present the system to stakeholders.
System are presented in different models which are created from different
perspectives. There are three perseptives such as external,behavior and structural.
Example of model are Framework, context diagram, Data Flow Diagrams (DFD) and
Entity Relationship Diagram (ERD). DFD are modelling the system from functional
aspects. It also can show the flow of data between systems.
While Entity Relationship Diagram are used to describe the relationship
between entities and attributes of entities. It widely available in database modelling.
Next, the another explaination will be available in this chapter.
17
3.6 FRAMEWORK
Framework is a basic structure that are needed to solve the complex problem
or as known as the tools and material or component. In the Diabetes Prediction
System, there are only two users that we called it as Admin and user.
For Admin, they need log into the system if they want manage their system.
After login, they are retrieved into their own interface (different interface with user‟s
interface) .They can add, delete or update the information segment. They also can
manage their profile, view prediction result of users, delete user and user‟s opinion
module. Admin also has right to add new admin for this system.
While for user, they need register firstly to gain user ID , email and password.
The user ID,user Name and password will be used by them to log into the system.
After successfully login, they can use Diabetes Prediction System by answer the
questionnaire that given. With the answer, the system will generate the result about
the user‟s potential to get Diabetes and they will advised to seek doctors to find out
real results. They also can view information about Diabetes and give their opinion
through „Contact Us‟ column.
18
Figure 3.6 :Framework for Diabetes Prediction System.
19
3.7 CONTEXT DIAGRAM
Figure 3.7 show the Context Diagram for Diabetes Prediction System. There
are two actor are involved in this system ; user and Admin. In context diagram, the
flow of the actors are explained and their ability in this system.
Figure 3.7 :Context Diagram
Description of Context Diagram
Based on figure 3.7, the DIABETES PREDICTION SYSTEM process at the
center of figure. There are two entities or actors are available are USER and ADMIN.
There are eleven data flows in the Context Diagram. Only two outgoing data flow
from ADMIN which consist of UPDATE INFORMATION and UPDATE
QUESTIONAIRRES. While from USER, only five outgoing data flow which consist
20
of SYSTEM EVALUATIONS ,LOGIN, REGISTER, PERSONAL DETAILS and
ANSWER OF QUESTIONAIRRES. For ingoing data flow, USER have only have
two which is DIABETES INFORMATION and RESULTS. ADMIN have only have
USER INFORMATION as ingoing data flow.
3.8 DATA FLOW DIAGRAM
Figure 3.8 show the Data Flow Diagram level 0 for the Diabetes Prediction
System. Since the figure 3.8 has been explained the flow of the actors; User and
Admin, in this chapter, the more details about the flow are explained with DFD
LEVEL 0 and following by DFD LEVEL 1. The functionality for each process also
will be described and able to help developer to understand their system.
21
Figure 3.8 Data Flow Diagram Level 0 for Diabetes Prediction System
22
Description of Data Flow Diagram level 0
There are two entities which are ADMIN and USER. While there are nine
processes are identified such as REGISTRATION, MANAGE USER, MANAGE
QUESTIONNAIRE, MANAGE RESULT, MANAGE SUGGESTION, MANAGE
RESPOND, MANAGE ADMIN, MANAGE SEGMENT INFO and lastly, REPORT.
Next, USER, QUESTIONAIRRE, RESULT, RECOMMENDATION, ADMIN,
DIABETES INFO, and RESPOND are the seven data stores for Diabetes Prediction
System.
1. USER input USER DETAILS into REGISTRATION which output is USER
DETAILS into USER data store.
2. USER input USER INFO into MANAGE USER which output is USER
INFO into USER data store.
3. USER input ANSWER DETAILS into MANAGE QUESTIONNAIRE which
output is ANSWER DETAIL into QUESTIONAIRRE data store and invoke
RESULT DETAILS input into MANAGE RESULT which output RESULT
DETAILS to RESULT data store and from MANAGE RESULT process
input RESULT PREDICTION into USER. While from RESULT datastore,
RESULT DETAIL is invoked into MANAGE SUGGESTION process where
will input the RECOMMENDATION data store. Then,
RECOMMENDATION data store will output RECOMMENDATION
DETAIL into MANAGE SUGGESTION process which output
RECOMMENDATION to USER.
4. USER input USER‟S RESPOND into MANAGE RESPOND which output is
USER‟S RESPOND into RESPOND data store.
5. ADMIN input ADMIN DETAILS into MANAGE ADMIN which output is
ADMIN DETAILS into ADMIN data store.
6. ADMIN input UPDATE QUESTION DETAIL into MANAGE
QUESTIONNAIRE which is output the QUESTION DETAIL into
QUESTIONNAIRE data store.
7. ADMIN input DIABETES INFORMATION into MANAGE INFO
SEGMENT which output is DIABETES INFORMATION into DIABETES
INFO data store.
23
8. All entities and data stores will input the REPORT into REPORT which is
output is REPORT .
3.9 DATA FLOW DIAGRAM LEVEL 1
3.9.1 Manage User
Figure 3.9.1: Data Flow Diagram Level 1 for Manage User
Description :
1. An USER input USER DETAIL to LOGIN process and then the process send
USER DETAILS into USER data store.
2. An USER input USER DETAIL to UPDATE USER DETAILprocess and then
the process send USER DETAILS into USER data store.
24
3. An USER input USER DETAIL to ADD USER DETAIL process and then the
process send USER DETAILS into USER data store.
4. An USER input USER DETAIL to DELETE USER DETAIL process and then
the process send USER DETAILS into USER data store.
3.9.2 Manage questionnaire for Admin
Figure 3.9.2: Data Flow Diagram Level 1 for Manage Prediction
Description :
1. An ADMIN input UPDATE QUESTION DETAIL to ADD QUESTION
process and then the process send QUESTION DETAILS into
QUESTIONNAIRE data store.
2. An ADMIN input UPDATE QUESTION DETAIL to UPDATE QUESTION
process and then the process send QUESTION DETAILS into
QUESTIONNAIRE data store.
3. An ADMIN input UPDATE QUESTION DETAIL to DELETE QUESTION
process and then the process send QUESTION DETAILS into
QUESTIONNAIRE data store.
25
3.9.3 Manage result
Figure 3.9.3: Data Flow Diagram Level 1 for Manage Result
Description :
1. An USER input HEALTH INFORMATION AND PERSONAL DETAIL to
ANSWER QUESTION process and then the process send HEALTH
INFORMATION AND PERSONAL DETAIL into QUESTIONNAIRE data
store.
2. An QUESTIONNAIRE data store input HEALTH INFORMATION AND
PERSONAL DETAIL into GET RESULT process and then ,the process
retrieve HEALTH INFORMATION AND PERSONAL DETAIL to RESULT
data store.
3. RESULT data store then input RESULT GENERATED into GET RESULT
process which is output RESULT to USER.
26
3.9.4 Manage Respond
Figure 3.9.4: Data Flow Diagram Level 1 for Manage Respond
Description :
1. An USER input USER‟S RESPOND to GIVE RESPOND process and then
the process send USER‟S RESPOND into RESPOND data store.
2. A RESPOND data store input USER‟S RESPOND into GET RESPOND
process and then ,the process send USER‟S RESPOND to ADMIN.
27
3.9.5 Manage Admin
Figure 3.9.5: Data Flow Diagram Level 1 for Manage Admin
Description :
1. An ADMIN input ADMIN DETAIL to ADD ADMIN process and then the
process send ADMIN DETAILS into ADMIN data store.
2. An ADMIN input ADMIN DETAIL to UPDATE ADMIN DETAILprocess
and then the process send ADMIN DETAILS into ADMIN data store.
3. An ADMIN input ADMIN DETAIL to DELETE ADMIN process and then
the process send ADMIN DETAILS into ADMIN data store.
28
3.9.6 Manage segment info
Figure 3.9.6: Data Flow Diagram Level 1 for Manage Segment Info
Description :
1. An ADMIN input INFO DETAIL to ADD SEGMENT INFO process and then
the process send INFO DETAILS into DIABETES INFO data store.
2. An ADMIN input INFO DETAIL to UPDATE SEGMENT INFO process and
then the process send INFO DETAILS into DIABETES INFO data store.
3. An ADMIN input INFO DETAIL to DELETE SEGMENT INFO process and
then the process send INFO DETAILS into DIABETES INFO data store.
29
3.10 ENTITY RELATIONSHIP DIAGRAM
Figure 3.10 show Entity Relationship Diagram of Diabetes Prediction System
(one to many) strong relationship
(one to many) weak relationship
An entity-relationship diagram (ERD) show that the entities information and entities
relationship. ERD is consist of identifying and defining the entities, determine
entities interaction and the cardinality of the relationship.
30
3.11 DATABASE MODELLING
Database are play important part in make sure the data and information in the
system display properly. Database are used to store the data.
Figure 3.11 The table that show the tables in the database Diabetes Prediction
System.
There are six table available in the database such as Admin, Feedback, Info,
Questionnaire, Result and user. For each table, there are attributes at every column.
31
3.11.1 Table Admin
Figure 3.11.1 : Table Admin
Table Admin contain AdminID, adminName, adminPassword, AdminEmail and
noPhone. In this table, Admin ID is a primary key and not null.
3.11.2 Table Respond
Figure 3.11.2 Table Feedback
There are only three attributes available such as feedbackID, email and feedback. For
this table, we have respondID as primary key while email is a foreign key which
reference from email attribute of user table.
32
3.11.3 Table Info
Figure 3.11.3 Table Info
There are only five attribute that consist of infoID, infoDetail , pic, adminID, and date.
InfoID is a primary key while adminID is a foreign key which is refer to admin table.
3.11.4 Table Questionnaire
Figure 3.11.4 Table Questionnaire
There are ten attributes that available in the table such as idquestionnaire, question,
answer1,answer2 , answer3, answer4, mark1, mark2, mark3, and mark4. In this table,
there is only primary key such as idquestionnaire
33
3.11.5 Table Result
Figure 3.11.5 Table Result
There are five attributes that available such as resultID, email, mark and risk. In this
table there are primary key and foreign key. The primary key is resultID while foreign
key is email which refer to table user.
3.11.6 Table User
Figure 3.11.6 Table User
There are five attributes that available such as email, password,firstName, lastName
and gender. In this table, the primary key is email and email is unique.
34
3.12 CONCLUSION
This chapter focused on the methodology of the software
development,requirements of software and hardware to achieve the objectives to build
the system. The system that will be build need able to be run and display on the
medium such as Google Chrome. With the right methodology that have been chosen,
the phases will able to followed correctly. The explaination of methodology, software
and hardware requirement has been described in this chapter.
Through this chapter, we will also got briefly how the system are modelling
into Framework, Data Flow Diagram, Context Diagram and Entity Relationship
Diagram. We also can understand how the flow of the system during design.
35
REFERENCES
1. Aiswarya Iyer, S. Jeyalatha and Ronak Sumbaly. 2015. Diagnosis Of Diabetes
Using Classification Mining Techniques. International Journal of Data Mining &
Knowledge Management Process (IJDKP) Vol.5, No.1.
2. S.Vijiyarani and S.Sudha. January 2013. Disease Prediction in Data Mining
Technique – A Survey. International Journal of Computer Applications & Information
Technology Vol. II, Issue I, January 2013 (ISSN: 2278-7720)
3. Laporan Tahunan Kementerian Kesihatan Malaysia, 2012.
4. American Diabetes Association. 2012.Standards of Medical Care in Diabetes
2012.
5. Galega officinalis. May 2009. Management Of Type 2 Diabetes Mellitus 4TH
Edition.
6. Illhoi Yoo , Patricia Alafaireet , Miroslav Marinov ,Keila Pena-Hernandez ,
Rajitha Gopidi ,Jia-Fu Chang and Lei Hua. 2012 . Data Mining in Healthcare and
Biomedicine: A Survey of the Literature. J Med Syst (2012) 36:2431–2448 DOI
10.1007/s10916-011-9710-5.
7. Data Mining - Rule Based Classification. 2017.
https://www.tutorialspoint.com/data_mining/dm_rbc.htm. Accessed on 12 February
2017.
8. Rashedur M. Rahman, Farhana Afroz., January 30th, 2013. Comparison of
Various Classification Techniques Using Different Data Mining Tools for Diabetes
Diagnosis. Department of Electrical Engineering and Computer Science, North South
University, Dhaka, Bangladesh.
36
9. What is Spiral Model? When to Use? Advantages & Disadvantages. 2017.
http://www.guru99.com/whatisspiralmodelwhentouseadvantagesdisadvantages.html.
Accessed on 28 March 2017.
10. Liu, H., Gegov, A. and Cocea, M. Granul. Comput. 2016. Rule-based systems:
a granular computing perspective. 1: 259. doi:10.1007/s41066-016-0021-6
37
APPENDIX A: GANTT CHART FYP1