DEVELOPMENT OF AN INTELLIGENT ACADEMIC
SUPPORTING TOOL
Liew Hui Chang
Bachelor of Engineering with Honors
(Electronic & Telecommunication Engineering)
2009
UNIVERSITI MALAYSIA SARAWAK
R13a
BORANG PENGESAHAN STATUS TESIS Judul: DEVELOPMENT OF AN INTELLIGENT ACADEMIC SUPPORTING TOOL
SESI PENGAJIAN: 2008/2009 Saya LIEW HUI CHANG
(HURUF BESAR)
mengaku membenarkan tesis * ini disimpan di Pusat Khidmat Maklumat Akademik, Universiti Malaysia Sarawak dengan syarat-syarat kegunaan seperti berikut:
1. Tesis adalah hakmilik Universiti Malaysia Sarawak. 2. Pusat Khidmat Maklumat Akademik, Universiti Malaysia Sarawak dibenarkan membuat salinan untuk
tujuan pengajian sahaja. 3. Membuat pendigitan untuk membangunkan Pangkalan Data Kandungan Tempatan. 4. Pusat Khidmat Maklumat Akademik, Universiti Malaysia Sarawak dibenarkan membuat salinan tesis ini
sebagai bahan pertukaran antara institusi pengajian tinggi. 5. ** Sila tandakan ( a ) di kotak yang berkenaan
SULIT (Mengandungi maklumat yang berdarjah keselamatan atau kepentingan Malaysia seperti yang termaktub di dalam AKTA RAHSIA RASMI 1972). TERHAD (Mengandungi maklumat TERHAD yang telah ditentukan oleh organisasi/ badan di mana penyelidikan dijalankan). a TIDAK TERHAD Disahkan oleh (TANDATANGAN PENULIS) (TANDATANGAN PENYELIA)
Alamat tetap: P.O.BOX 116 94000 BAU, SARAWAK MR. TAY KAI MENG Nama Penyelia
Tarikh: 8 April 2009 Tarikh: 8 April 2009
CATATAN * Tesis dimaksudkan sebagai tesis bagi Ijazah Doktor Falsafah, Sarjana dan Sarjana Muda. ** Jika tesis ini SULIT atau TERHAD, sila lampirkan surat daripada pihak berkuasa/organisasi
berkenaan dengan menyatakan sekali sebab dan tempoh tesis ini perlu dikelaskan sebagai SULIT dan TERHAD.
This Final Year Project attached here:
Title : Development of an Intelligent Academic Supporting Tool
Student Name : Liew Hui Chang
Matric No : 14356
has been read and approved by:
__________________ _________________
Mr. Tay Kai Meng Date
(Supervisor)
DEVELOPMENT OF AN INTELLIGENT ACADEMIC SUPPORTING TOOL
LIEW HUI CHANG
Thesis is submitted to
Faculty of Engineering, Universiti Malaysia Sarawak
in partial fulfillment of the requirements
for the degree of Bachelor of Engineering
with Honours (Electronic and Telecommunications Engineering) 2009
Dedicated to my beloved family and friends
i
ACKNOWLEDGEMENT
First and foremost, I would like to thank my supervisor, Mr Tay Kai Meng for
his insightful guidance and encouragement throughout this work. Without his advice,
I would not have been able to proceed and bring the project to a satisfactory
completion. My appreciation also goes to my mentor, Ms Shafrida Sahrani for her
constant motivation and support.
I would like to express my gratitude to my parents and siblings for their
continuous support and encouragement in my study over the years. Besides, I would
like to thank all my friends especially Salex, Mee Lee, Tze Ling, Kok Cheow, Yee
Chiang, Kim Khoon and Kian Tat for their supports and companion during four
years study in UNIMAS. My special thanks go to Ming Shinn, for his care and
continuous moral support.
Last but not least, I wish to thank Dr Darrien Mah and all the staff of Faculty of
Engineering and also those who directly or indirectly help in this project success.
vi
TABLE OF CONTENTS
Page
ACKNOWLEDGEMENT i
ABSTRACT ii
ABSTRAK iv
TABLE OF CONTENTS vi
LIST OF TABLES x
LIST OF FIGURES xii
ABBREVIATION xv
CHAPTER 1 INTRODUCTION
1.1 Overview 1
1.2 Problem Statements 2
1.3 Objectives 3
1.4 Project Overview 4
1.4.1 Electronic Notification System 4
1.4.2 ETree based Automated Document
Cluster System
5
1.5 Thesis Outline 6
CHAPTER 2 LITERATURE REVIEW
2.1 Introduction 8
vii
2.2 Electronic Notification (E-notice) System 9
2.3 Short Message Service (SMS) messaging 11
2.4 Send SMS messages from a computer 14
2.5 Software Development Kit 15
2.6 Data Clustering and Visualization Techniques 15
2.6.1 Self-Organizing Maps 17
2.6.2 Self-Organizing Algorithm 19
2.6.2.1 Initializing the Weights 19
2.6.2.2 Get Best Matching Unit
(BMU)
19
2.6.2.3 SOM Training 20
2.6.3 SOM Visualization 21
2.6.4 Growing Self-Organizing Maps
(GSOM)
21
2.6.5 Evolving Tree as a type of Growing
Self-Organizing Maps
22
2.6.5.1 Operation of Evolving Tree
(ETree)
23
2.7 Documents Classification, Visualization and
Organization
25
2.7.1 The Self-Organizing Map Approach 26
2.7.2 The Growing Hierarchical Self-
Organizing Map Approach
29
2.8 Summary 31
viii
CHAPTER 3 METHODOLOGY
3.1 Introduction 33
3.2 System Development Procedures 33
3.3 Motivation of Invention for Electronic
Notification System
35
3.3.1 Overall Operation 37
3.3.2 Terminal Program 39
3.4 Motivation of Invention for ETree based
Automated Document Clustering System
40
3.4.1 The Proposed ETree based Automated
Document Clustering System
41
3.4.2 Abstract Content Similarities
Matching
43
3.4.3 Construction Procedure of the Proposed System
46
3.4.3.1 Construction of Document
Matrix
47
3.4.3.2 Construction of ETree 51
3.4.4 Development of Graphical User
interface and Visualization
57
CHAPTER 4 RESULTS, ANALYSIS AND DISCUSSION
4.1 Introduction 59
4.2 The Electronic Notification System 59
ix
4.2.1 The Designed Graphical User
Interface
59
4.2.2 Evaluation of the System 67
4.3 The ETree Automated Document Clustering
System
68
4.3.1 The Designed Graphical User
Interface
68
4.3.2 Evaluation of the System 70
4.3.2.1 50% of Similarity Threshold 72
4.3.2.2 40% of Similarity Threshold 79
4.3.2.3 30% of Similarity Threshold 82
4.3.2.4 20% of Similarity Threshold 87
4.3.3 Further Evaluation 90
4.3.4 Further System Accuracy Evaluation 95
4.4 Summary 109
CHAPTER 5 CONCLUSIONS & RECOMMENDATIONS
5.1 Conclusions 110
5.2 Project Limitations 111
5.3 Future Works 112
REFERENCE 113
APPENDIX A 119
xv
ABBREVIATION
2-D Two-Dimensional
BMU Best Matching Unit
BSS Base Station System
CDMA Code Division Multiple Access
ENCON Engineering Conference
ESOM Evolving Self-Organizing Map
ETREE Evolving Tree
GHSOM Growing Hierarchical Self-Organizing Map
GMSC Gateway Mobile Switching Center
GSM Global System for Mobile Communications
GSOM Growing Self-Organizing Map
GUI Graphical User Interface
HLR Home Location Register
HTML HyperText Markup Language
IDE Integrated Development Environment
PBS Personal Bibliographic Software
SDK Software Development Kit
SIM Subscriber Identity Module
SMC Short Message Center
SME Short Message Entity
SMS Short Message Service
xvi
SOM Self-Organizing Map
TDMA Time Division Multiple Access
UNIMAS University Malaysia Sarawak
VLR Visitor Location Register
xii
LIST OF FIGURES
Page
2.1 Typical organization of network elements in a GSM
network supporting SMS [Adapted from [18]]
12
2.2 Basic Evolving Tree [Adapted from [35]] 23
2.3 Overview of the construction and operation of the
WEBSOM system [Adapted from [41]]
28
2.4 The schematic diagram of the proposed approach
[Adapted from[32]]
30
3.1 Activities and steps use to develop the academic
supporting system [Adapted from [46]]
34
3.2 Simplex communication 35
3.3 Illustration on the operation of the E-notification system 38
3.4 Point-to-multipoint messaging 39
3.5 Flow chart for overall process for the terminal program 40
3.6 The general schematic diagram of the proposed approach 43
3.7 Keywords matching process 43
3.8 Sequences of keywords in document matrix form 44
3.9 Keywords matching process 45
3.10 Construction procedure of the ETree system 46
3.11 Flow chart for the textual-preprocessing phase 47
3.12 Stop words elimination 48
xiii
3.13 Abstract content from the raw patent technical article 50
3.14 Textual-preprocessing 50
3.15 Document matrix 50
3.16 Split node and its children nodes 51
3.17 Trunk node and leaf nodes 52
3.18 Fundamental operations of Evolving Tree 54
3.19 Best matching unit 55
3.20 Finding BMU for two nodes which owned the same
matching value
56
3.21 Overall flow chart for ETree construction phase 57
3.22 Visualization for ETree based Automated Document
Clustering System
58
4.1 Login Form 60
4.2 Main Menu Form 60
4.3 Students information selection based on year of study
form
61
4.4 Students information selection based on courses form 61
4.5 Students information based on program selection form 62
4.6 Students information management form 62
4.7 Short message service based on year of study form 63
4.8 Short message service selection based on courses form 63
4.9 Short message service selection based on program 64
4.10 Short message service management 64
4.11 Change login password form 65
xiv
4.12 System exit form 65
4.13 Introduction form 69
4.14 Main Menu form 69
4.15 Search form 70
4.16 Training result of ETree system for 50% of similarity
threshold
74
4.17 ETree Map for 50% of similarity threshold 75
4.18 Summary in pie chart for result obtained in 50% of
similarity threshold
78
4.19 Training result of ETree for 40% of similarity threshold 79
4.20 ETree Map for 40% of similarity threshold 80
4.21 Summary in pie chart for result obtained in 40% of
similarity threshold
82
4.22 Training result of ETree for 30% of similarity threshold 83
4.23 ETree Map for 30% of similarity threshold 84
4.24 Summary in pie chart for result obtained in 30% of
similarity threshold
86
4.25 Training result of ETree for 20% of similarity threshold 87
4.26 ETree Map for 20% of similarity threshold 88
4.27 Summary in pie chart for result obtained in 20% of
similarity threshold
90
4.28 System accuracy comparison 95
4.29 System accuracy versus similarity threshold for 100, 150,
265 articles.
109
x
LIST OF TABLES
Page
3.1 Trained result 52
3.2 Matching result for Figure 3.18 55
3.3 Calculation to Find the BMU for two nodes which
owned the same matching value
56
4.1 List of the 50 selected ENCON 2008 articles 71
4.2 Conference articles and their cluster with 50% of
similarity threshold
76
4.3 Conference articles and their cluster with 40% of
similarity threshold
80
4.4 Conference articles and their cluster with 30% of
similarity threshold
84
4.5 Conference articles and their cluster with 20% of
similarity threshold
88
4.6 Validation result for 50% of precision 91
4.7 Validation result for 40% of precision 92
4.8 Validation result for 30% of precision 93
4.9 Validation result for 20% of precision 94
4.10 Overall system accuracy 94
4.11 List of the 100 selected ENCON 2008 articles 96
4.12 List of the 150 selected ENCON 2008 articles 98
xi
4.13 List of the 265 full set of ENCON 2008 articles 102
4.14 System accuracy 108
ii
ABSTRACT
Academic supporting tool is defined as an educational development services
which is developed as a guide to universities and colleges in order to support and
help in academic progress. In this thesis, two academic supporting tools; e.g., an
electronic notification system, and an Evolving Tree (ETree) based automated
documents clustering system are developed, and evaluated. The electronic
notification system with Short Message Service (SMS) technology is developed to
facilitate the communication among lecturers and students. It can be viewed as a
mobile, efficient and environment-friendly communication platform among lecturers
and students. It allows latest news, general notices, instant class cancellation alerts
and etc. to be delivered to students instantly, through SMS. The electronic
notification system is developed using Visual Basic language. Practicality of the
developed system is further evaluated with the help of a lecturer in UNIMAS.
Positive results are obtained. The ETree based automated documents clustering
system is developed to allow conference papers to be clustered, organized and
visualized based on their abstract similarity. Classification of conference papers is
always a tedious, time-consuming and expensive task, to be carried out by
experience academicians. The developed system is a decision supporting tool for
conference organizers and technical committees. It eases the task of classification of
conference papers. With regards to the study of neural network, ETree is proposed
as an improvement of Self Organizing Map (SOM). As compared with conventional
SOM, ETree does not require map size to be predetermined. From literature, the use
of SOM in document clustering, organization and visualization problems has been
iii
widely reported. However, from literature review, the main application of ETree
centered at image processing, and no report on the use of ETree in document
clustering, organization and visualization problems have been found. To some extent,
this piece of work can be viewed as an earlier investigation on the use ETree for
document classification problems. The ETree based automated documents
clustering system is evaluated with information from Engineering Conference
(ENCON) 2008. Correctness of the clustering result is further evaluated with the
help of a Technical Committee from ENCON 2008. Experiments show that it is a
potential system, and positive results are obtained.
iv
ABSTRAK
Academic Supporting Tool adalah satu perkhidmatan pembangunan pendidikan
yang dicipta sebagai satu panduan untuk menyokong dan membantu universiti dan
kolej dalam kemajuan ilmiah. Dalam tesis ini, dua sistem penyokong akademik iaitu
sistem pemberitahuan secara elektronik dan sistem klasifikasi bagi dokumen secara
automatik berdasarkan algoritma Evolving Tree (ETree) telah dihasilkan dan dinilai.
Sistem pemberitahuan secara elektronik dengan menggunakan teknologi Khidmat
Pesanan Rungkas dihasilkan untuk memudahkan komunikasi antara pensyarah dan
pelajar. Sistem ini merupakan satu teknologi mudah alih, cekap dan mesra untuk
meningkatkan komunikasi antara pensyarah dan pelajar. Ini membolehkan berita
terkini, notis am, notis pembatalan kelas dan sebagainya diumumkan kepada pelajar
melalui perkhidmatan SMS. Sistem pemberitahuan secara elektronik ini telah
dihasilkan dengan menggunakan perisian Visual Basic. Sistem ini telah dinilai dari
segi praktikal dengan bantuan seorang pensyarah di UNIMAS. Maklum balas positif
telah diperolehi melalui penilaian ini. Sistem klasifikasi bagi dokumen secara
automatik telah dihasilkan untuk membolehkan kertas persidangan diklasifikasikan
dan disusun berdasarkan persamaan abstrak dokumen. Klasifikasi bagi kertas
persidangan merupakan satu tugas yang agak membosankan, mengambil masa,
memerlukan kos yang tinggi dan perlu dilaksanakan oleh ahli-ahli akademik yang
berpengalaman. Sistem yang dihasilkan merupakan satu alat sokongan dalam
membantu pihak penganjur dan jawatankuasa persidangan membuat keputusan. Ia
juga meringankan tugas pengelasan kertas persidangan. Berdasarkan pengajian
dalam bidang Neural Network, ETree diperkenalkan sebagai satu peningkatan bagi
v
Self-Organizing Map (SOM). Berbanding dengan konvensional SOM, ETree tidak
perlu menentukan saiz peta terlebih dahulu. Dari kajian literasi, penggunaan SOM
dalam pengelasan, organisasi dan pembayangan dokumen telah dinyatakan secara
meluas. Bagaimanapun, daripada kajian literatur, aplikasi ETree hanya tertumpu
pada pemprosesan imej. Manakala tiada laporan mengenai penggunaan ETree dalam
pengelasan, penyusunan dan pembayangan dokumen dinyatakan. Kajian ini boleh
dikatakan sebagai satu pertemuan baru bagi penggunaan ETree dalam pengelasan
dokumen. Sistem ini telah dinilai dengan menggunakan maklumat yang diperolehi
daripada Engineering Conference (ENCON) 2008 yang dianjurkan oleh UNIMAS.
Ketepatan keputusan sistem telah dinilai dengan bantuan seorang ahli teknikal
jawatankuasa daripada ENCON 2008. Keputusan eksperimen menunjukkan bahawa
ia merupakan satu sistem yang berpotensi dan maklum balas positif telah diperolehi
melalui penilaian ini.
1
CHAPTER 1
INTRODUCTION
1.1 Overview
An academic supporting tool becomes very important and popular nowadays
with respond to internal and external pressures for accountability in higher education.
An academic supporting tool is an educational development services which is
developed as a guide to universities and colleges in order to support and help in
academic progress.
One of the popular existing academic supporting tools is the Endnote [1].
Endnote is a Personal Bibliographic Software (PBS) package designed to manage
bibliographic references to all kinds of information such as printed sources, non-
printed materials, electronic resources and etc [1]. It is used for managing references
for those involved in teaching, research and supervising postgraduate research
students. Bournemouth is one of the universities who purchased a three years site
license for Endnote [2]. Another example is Morpheus, an academic support system
at University Malaysia Sarawak [3]. It is developed to support students’ and
academicians’ teaching-learning process. Morpheus is a free learning management
system that enables you to create powerful, flexible, and engaging online learning
experiences. It is designed to support a learning style, known to educators as Social
2
Constructive Pedagogy. Morpheus enables you to add static course materials,
interactive course materials as well as other social course materials [3].
The purpose of this project is to develop an environment-friendly system to
support and help in academic progress.
1.2 Problem Statements
Conventional notice board in a university campus is a paper-based offline
notification system which allows lectures to display messages and announcements to
their students by posted printed or written information [4]. Problems of this
approach arise when students cannot receive the instant news like classes have been
removed or cancelled at the last moment on time. Therefore the electronic
notification system is proposed to provide an efficient and environment-friendly
platform to enhance communication between lecturers and students.
Classification of conference papers is a tedious, yet important task for
conference organizers, as the number of technical papers increases. With a growing
amount of conference paper, there is a need to develop a computer aided system to
support this task [5]. In this project, an Evolving Tree (ETree) based supporting
system is developed to cluster conference papers based on their abstract similarity.
The ETree is then visualized in two-dimensional (2-D) similarity plot.
3
1.3 Objectives
The main objective of this project is to develop an environment-friendly system
to support and help in academic progress. In order to achieve this, several objectives
must be reached:
The objectives of this project as follow:
i. To develop an Electronic Notification System that able to provide efficient
and environment-friendly platform to enhance communication between the
lectures and students.
ii. To develop an Evolving Tree (ETree) based Automated Document Clustering
system to cluster conferences paper based on their abstract similarity.
iii. To visualize the ETree in two-dimensional similarity plot.
iv. To investigate on the use of ETree technique in document clustering problem.
Top Related