Cmputing: : Engaging Communities

Patterns Using MEMISP ... Enhancement of Parallel Thinning Algorithm ... Complexity of Java Programming Assignment...

0 downloads 58 Views 2MB Size
r,

ýýý

r



'ýJ

ý2

C-

ww

"J

VJ

ttp

://w

,_.

-ML

iGLJa.

.pd fsp

'JIL

litm

erg er. co m

)

11

ý1ý ý,ilýn iii

th

ni

(V isi

,

De mo

.-ý.

v

ia2OOV

Sociai Cmputing: Communities Engaging : L-,

ý-ý .6, -''

Küching?Sarawak ff

Organised

by:

In Collaboration

A.

7

with:

InfUnmitiun & Communications Technology Unit UCT) Chef Min , ter s Deparirnent of Sm awak

'a"" Global Information & Tuleeommunleatlon Institute

FIFTH CONFERENCE

INTERNATIONAL

ON

TECHNOLOGY

INFORMATION

IN ASIA

htt p:/

/w w

w.

pd fsp

litm e

rge

r.c

om

)

2007

De

mo

(V

isi t

PROCEEDINGS OF CITA'07

Edited by Alvin W. Yco Jane Labadin Wang Yin Chai Tan Chong Eng

Organised bY Faculty

Kuching,

of Computer

Sarawak,

Science & Information

Malaysia

Technology,

Universiti

Malaysia

Sarawak

July 09-12,2007

) er. co m erg litm fsp .pd ttp ://w ww th (V isi mo De

Proceedings of CITA'07 The Fifth International Conference on Information Technology in Asia July 09-12,2007, Kuching, Sarawak, Malaysia

ISBN #: 983-9257-66-8

Universiti Malaysia Sarawak 94300 Kota Samarahan Sarawak Malaysia

Copyright © 2007 by Universiti Malaysia Sarawak

Editorial Preface This is the Proceedings of the 5'h International Conference on Information Technology in Asia (CITA'07), held between 91h 12`h July 2007 in Kuching Malaysia. CTTA'07 is organised by Universiti Malaysia Sarawak in collaboration with the ICT Unit, Chief Minister's Department of Sarawak, and Global Information and Telecommunication Institute. The conference provides an important platform to showcase, in particular, state-of-the-art Technologies Information Communication in (ICTs) in Asia of and applications and research while promoting the exchange of ideas and research results with researchers all over the world. The theme of CITA'07 on Social Computing: Engaging Communities underlies our efforts to highlight emerging trends and technologies. This theme not only covers the traditional roles ICTs In but lives, in contemporary ones well. also as addition, the new technological play our paradigm encompasses not only the technological aspects but also the social aspects of computing, bringing people and communities closer together in their work and play.

isi th

ttp

://w

ww

.pd

fsp

litm

erg

er. c

om

)

This year, we have received a good response of 166 submissions from 10 countries. An International Review Committee reviewed these submissions, and from these, 46 full papers and 22 short papers have been included in this conference proceedings. These papers are presented under 10 major tracks which includes: Agents and Autonomous Systems, Communications Technologies. Community Informatics, Computational Models and Systems, Education Technologies. Emerging Technologies and Platforms, Image Processing and Visualisation, Information Systems. Knowledge Networks and Management, and Software Engineering. This diverse yet complementary range of topics cover the omnipresent nature of ICTs applied in all aspects of our lives.

De

mo

(V

It is hoped that this conference will also provide a platform to bring together researchers and practitioners to share their knowledge and experiences in preparing us in staying ahead in today's challenging and dynamic environment. We would like to acknowledge and express our gratitude to the many people who have contributed greatly to the conference. I would like to thank the members of the International Review Committee for reviewing the papers. and the members of the organising committee for their tireless effort in making this conference a success. We extend our sincere appreciation to all sponsors for their generous contribution. We wish you all an enjoyable conference with fruitful Datang" to our visitors to the Land of the I lornbills.

Dr. Alvin W. Yeo Conference Chair

iii

deliberations and hid a warm " Selamat

International Review Committee Members Sadeghi Ahmad-Reza Bochum, Germany Ruhr-University W. Yeo Alvin Universiti Malaysia Sarawak, Malaysia Andrew

Walton

Imperial College London, United Kingdom Bettina Berendt Humboldt-Universität zu Berlin, Germany Bong Chin Wei Wawasan Open University, Malaysia Chen Chwen Jen Universiti Christian

Malaysia Guetl

Sarawak,

Malaysia

Gra: University of Technology, Austria Daniel Cunliffe University of Glamorgan, United Kingdom Douglas Chai

Universiti Jane

Malaysia

Sarawak,

Malaysia

er. c

erg litm fsp .pd ww ://w th ttp isi (V mo

Universili

De

Universiti Malaysia Sarawak, Malaysia Farid Meziane University ofSalford, United Kingdom Gary Marsden University of Cape Town, South Africa Geoff Holmes University of Waikato, New Zealand Hong Kian Sam

om )

Australia

Edith Cowan University, Mit Edwin

Stewart

Universiti Malaysia Sarawak, Malaysia Lau Bee Theng Swinhurne University of Technology, Malaysia Law Choi Look Nanyang Technological University, Singapore Lim Chee Peng Sains Malaysia,

Masood

Masoodian

University Mazian

of Waikato, Abas

Narayanan

Kulathuramaiyer Malaysia Sarawak,

Universiti

of Technology,

Australia

Universiti Sains Malaysia, Malaysia Tan Chong Eng Universiti Malaysia Sarawak, Malaysia Teh Chee Siong Universiti Thomas

Malaysia

Sarawak,

Malaysia

Mandl

University of Hildesheim, Germany Wang Yin Chai Universiti Malaysia Sarawak, Malaysia Wong Chee Weng

MIMOS Berhad, Malaysia Md. Yazid Mohd Saman Terengganu,

University

T. Ramayah

New Zealand

Malaysia

Malaysia

Marshall

Swinburne

Malaysia

Universiti

Putra Malaysia,

The University of West Indies, Barbados Sinnappan Sukunesan

Labadin

Universiti

Noordin Ahmad Geoinfo Services Sdn Bhd, Malaysia Pat Hall Open University, United Kingdom Patricia Anthony Universiti Malaysia Sabah, Malaysia Patrick Then Swinburne University of Technology, Malaysia Paula Bourges Waldegg Research and Advanced Studies Center (C/N VESTAV), Mexico Reinhard Schaler University of Limerick, Ireland Robert H. Barbour Unitec, New Zealand Rosziati Ibrahim Universiti Tun Hussein Onn, Malaysia Sally Jo Cunningham University of Waikato, New Zealand Seng Wai Loke La Trobe University, Australia Shahren Ahmad Zaidi Adruce Universiti Malaysia Sarawak, Malaysia Shapiee Abdul Rahman Universiti Malaysia Sarawak, Malaysia Shyamala C. Doraisamy

Malaysia

Universiti Malaysia

b Bolhassan Alamshah Universiti Malaysia Sarawak, Malaysia

Noor

iv

Malaysia

Sarawak,

Malaysia

Organising Committee

Conference Advisors Abdul Rashid Abdullah Khairuddin Ab Hamid Narayanan Kulathuramaiyer Johari Abdullah Conference Chairman Alvin W. Yeo Conference Co-chairs Noor Alamshah ßolhassan Jane Labadin Secretary Sharin I lazlin Iluspi Sze San Nah

De

mo

(V

isi

th

ttp :

//w

ww

.pd

fsp

litm

erg er. co m

)

Programme Committee Wang Yin Chai (Chairperson) Tan Chong Eng Edwin Mit Shapiec Abd Rahman Wong Chet Wong Terrin Lim

V

Organising Committee Publication Committee Suhaila Saee (Chairperson) Eaqerzilla Phang Yanti Rosmunie Bujang Rosita Mohamed Othman

Technical Committee Muhammad Asyraf Khairuddin (Chairperson) Ahmad Hadinata Fauzi Lee Jun Choi Sarina bt Ahmad Sharipah Norainie Said Akil Zulkifli b Ahmat

Logistics Committee Lau Sei Ping (Chairperson) Mohamad Nazri b Khairuddin Tiong Wei King

Wan Hosen Wan Shekh Kamal Sarkawi

Phang Pian

Razeki b Jelihi Mohd Arif Ibrahim Zulhilmi Roslan Hairullizam Azemi Nurul Hartini bt Minhat Elizabeth Suni

Sarah Flora Samson Juan Nuha Loling Othman Protocol Committee Mohd Johan Ahmad Khiri (Chairperson) Seleviawati Tarmizi Inson Din Nor Surayati Mohd Usop Noor Hazlini Borhan Nur Khairina Abdullah Doris Francis Harris Zuraini Ramli Muhsin Apong

Nur Rabizah bt Adeni

De

Publicity Committee Hamimah Ujir (Chairperson) Sy Fazlin Seyed Fadzir Ling Yeong Tyng Roziah MajIan

mo (

Vi sit

htt p:/ /w

ww .pd

fsp litm erg

er.

co

m)

Workshop Committee Abdul Rahman b. Mat (Chairperson) Stephanie Chua Hui Li Wang Hui Hui

Finance Committee Hadijah Morni (Chairperson) Hamisah Ahmad Selina Jawawi Sponsorship Committee Irwandi Hipni Mohd Hipiny Fatihah Ramli Vanessa Wee Bui Lin

Conference Management Toolkit Committee Chiew Kang Leng (Chairperson) Terrin Lim

(Chairperson)

Multimedia Presentation Committee Nurfauza Jali (Chairperson) Mohd Imran Bandan Syahrul Nizam Junaini Jonathan Sidi

Website Design Committee Nurfauza Jali (Chairperson) Nurul Zawiyah Mohamad

VI

3"Us11 Khid, nVt Niahº., niat 4kadrt, n: WWVhIt. )l'1'l N41t. Iºt'; alA SiAKAWAiº

Table of Contents FULL

PAPER

Communication Technologies Agents and Autonomous Systems

A Novel Mobility Prediction in 802.11 Infrastructure Networks by Location Tracking and Data Mining

Adaptive Threshold for Lane Detection with Hough Transform Azali Saudi, Jason Ahmad Hijazi

Teo & Mohd

Biyu Issac, Khairuddhin Tan

Hanafi

Community Informatics System

)'ogila Kanesin & Gian ('hand

A. V. Hail, Thapa

Telecentre Context

6

Ganesh

B. Ghimire, Prakash AManandhar &

Evaluation

the

Norsu: ila YU acoh, Afardina Ahdullah

Malaysian

Roger W. Harris

De mo

(V

isi

th

ttp

://w

ww

i,

Knowledge Networks and Management

23

Shti"umala Doraisunn"

Ych

7/

Rhumala Ahmed & Jalil Md. Desa .

-x

K.

Simplified Reducing Network

36

84

Network

Amite hhu ('hukruhurly, Al. Ahdur Ru_: uk, \a/L'es A1un oor K Al. Ahdus Sohhun . .

28

& ('hung-//sing

&

Kamm. hia_nah Abdul Samad Ismail, Stephan Olariu, Muhd A=aini Afaaru/ B& Shaharuddin Salleh . .

Handover Algorithm for Multiple Access Enabled Mobile Devices

Comparing Normalization Procedures in Multiattribute Decision Making Under Various Problem Settings Suhratu ('hukrahorti .

Ismail

Sink Node Mobility in Covering Monitoring Area for Data Collection in WSNs

Noor Alan Subramaniam

Precision Improvement using Contextual Retrieval and Data Fusion Approaches A-- A_rinudin Alidin, c& FaNmah Ahmad

Mahamod

Optimizing the Rebroadcast of Ad-hoc OnDemand Distance Vector (AODV) Routing Protocol

Using Fuzzy Integral to Evaluate the Webbased Application Ha_ura Zulrulil, Ahd: d A_im Ahd Ghani. Mohd Hasan Selamat & Ramlan ifahmod

64

Model Validation for Total Electron Content (TEC) Over an Equatorial Station

ii in

58

Architecture Management and Handoff Protocol for Next-Generation Wireless Systems 4: ila Lahr Yuso/, Mahamod Ismail & . Norhahiah Micran

r.c om )

Patrick Ishwor

Sodhti .

.pd fsp litm erg e

Dhakhwa,

51

Using Economic Model to Improve the Performance of BHR Dynamic Replication Algorithm

HCI for Non-literates - An Experience from Sambad Project Sagun

43

Finding Granular Features using Rough-PSO in IDS Ana=ida Zoinal, Mohd Aizaini Maarn/'& Siti Marivam Shamsuddin

1

Benefits of In-House e-Procurement Norhidavah Mohamad & Wan Khairuz_aman Wan Ismail

Ah Hamid & C. E.

Xv

for XOR Based Approach Broadcast Redundancy with East in Wireless Coverage Ad-Hoc

Networks

Aid.. hunured Islam, Aid Abu/ Ahaer, A1d. Va rul (slum & Aid. Afuhidul /slams

vii

94

Power and Congestion Optimized OnDemand Multi-path Routing (PCOOMR) in Mobile Ad-hoc Networks S. Amirtharai & K. Muneeswaran Energy Efficient Algorithm for Scatternet Sensor in Bluetooth-Based Formation Networks Sumair Khan & Kashif Zia Analysis Delay Performance of Two SlotBased Contention Resolution Algorithms Chairat Srichavengsup, Warakorn Wuttisittiku/kij Nitiyarom & Lunchakorn

Computational Models and Systems 101

Modified Sequencing using Operation Particle Swarm Optimization Zalmiyah Zakaria & Safaai Deris Heuristic-based Self-Organizing Initialisation for Learning Vector Quantisation (LVQ) Md. Sarwar Zahan Tapan & Chee Siong Teh

106

112

Education Technologies E-Portfolio Using to Supplement the Teaching of Mandarin as a Foreign Language Goh Ying Soon, Mohamad Zain

Tey Meng

Kar

& Jasni

161

167

CSG Tree: Evaluation on Performance between Solid Modelers Awanis BI Romli, Habibollah Haron & Abdul Syukor Mohamad Jaya

173

Comparison of FCM and FCV Clustering in Cluster-Based Compound Approach Selection & Sinarwati Huspi Hazlin Sharin Mohamad Suhaili

178

117

Emerging Technologies and Platform Environment for Knowledge Acquisition Bridging the Semantic Gap Narayanan Chin & Lim Phei Kulathuramaiyer

124

litm

erg e

r.c

om

)

Factors Influencing the Adoption of eLearning as a Teaching Tool: A Case of Universiti Utara Malaysia Hafi_ah Mohamud Hshollah & Kamil Md. ldris

Awang

mo

(V

isi t

htt p

://w w

w.

Dayang Rohaya Dhany Saputra, Rambli & Foong Oi Mean

De

An Efficient Data Structure for General TreeLike Framework in Mining Sequential Patterns Using MEMISP

& Alvin

in Anaphora Challenges News Media Genre P. Nund

190

Knowledge Machines: A New Computers Syed V. Ahamed

/40

Breed of 197

Enhancement of Parallel Thinning Algorithm for Handwritten Characters Using Neural Network

145

Adeline Narayanan

& Habibollah

Engkamat

Haron

203

Parametric Approach A Novel to Define for Motion Lips Equation Polynomial Estimation

151

W. Yeo Resolution

Web

Image Processing and Visualisation

Contextual Sense Disambiguation Ong,

Semantic

133

Enhanced Term-Specific Smoothing Using Document Word Clusters Rashid U& Bhatti M. A. for Knowledge Groupware Enhancing Management Yu-N Cheah Chin Siou Kulathuramaiver

Data Mediation over Framework Nor Adnan Yahaya

pd fsp

Knowledge Networks and Management

184

in the

Siti Salwa Salleh, Rahmita Wirza Rahmat,

156

Ramlan Mahmod Reduction

& Fatimah

of Artifacts

Ahmad

and Edge Preservation

of Images K. Sivakami Sundari & V. Sadasivam

viii

210

216

Photo Based Walkthrough and Caching Strategies

Yew Kwang Hooi, Mohamed Zakaria & Lai Chen Yes

One Dimensional Algorithm Douglas Chai

Pre-Fetch

with

Nordin

SHORT PAPER

h.

222

Barcode

Agents and Autonomous

Reading

Systems

Intelligent Agent for Online C Programming Assignment Assistance System

230

Syra G. Mokunjil

& Patricia

Anthony

283

Human Computer Interaction Building Trustworthy e-Commerce Ah Razak Che Hussin

Community Informatics

Website

236

Development of Modular Motion Sensing Game Controller H. G. Rodney Tan, Y.L. Jacob Tan & V H. Mok

ICT Implementation Issues in Local Government: A Conceptual Framework Rugavah Hj. Hashim. Rosmimah Mohd. Roslin & Adnan Jamaludin

242

A Cost-Effective, Customizable DTMFBased Interactive Voice Response System Muhammad Inamullah. M Qasim RaJiq & Ekram Khan

286

Communication Technologies 24'

Lifting 253

Modeling Variable Transmission Control Data Gathering Trees for Wireless Sensor Networks Lim Jong C'hern & K. Daniel Wong

290

Upgrading Performance of Ad-hoc OnDemand Distance Vector (AODV) Routing Protocol in Mobile Ad Hoc Networks Noor Alan Ahmad, Shamala K. Suhramaniam & Jalil Md. Desa

295

co m)

Visual Feedback Cues On Virtual Simulation A.A. Faieza & N. Ismail

pli tm

erg

er.

Software Engineering

w. /w w p:/ htt Vi sit

Binti Md. Rejah & Ruhaidu

Binti

mo (

Muwarnv Romli

pd fs

Automating the Process of Measuring the Complexity Java Programming of Assignment 259

De

Phichet 41oungnoul, , A'ohchui Dejhun

A Fine Grained Location Modeling for RoleBased Access Control Muhammad

Naheel Tahir

Laboratory Door (LDMS) Prototype ADAMView

Arunee A'unwihoon

c&

300

SMS Managament System for Direct Sales and Network Marketing

265

Monitoring System using ADAM and

Hu: ulilu finti Aumuludin, HanaVanti Binh L/aJit, A'hairu/ Bin Muhd Sukri, Amin . /.srecfra Rahmi Binti A. Humid A Lu Chin Leung

Packet Loss Recovery in Media Specific FEC. Audio Transmission by Least Square Method

{'oo Nvuk Ater & Ali Selamat .

303

Enhanced MAC An Protocol for Differentiated Services in Mobile Ad-hoc Networks (MANET) Profile-Based -A Approach R. (: zuwsrkurun

2'0

A. 1'. Rhymrnd

Education

(Ilhuriuruj

306

Technologies

Information Systems Relational Phones

Promoting Classroom Participation Interaction Using Wi-Fi Technology

Expressive Queries on Mobile

Ruhrcu Ahmud & Sumeem Ahdu! Auree'm .

Ahmed llishunt Luinul Ahd Rauh K Gan ('hin

''h

ix

Ahidin. Lur

and

Ahd lludi

30V

Image Processing and Visualisation

Knowledge Networks and Management Classification of Students' Performance in Programming Course using Decision Tree Cik Fazilah Hibadullah & Norita Md Norwawi Approach for A Hybrid Classification Content-based Email Filtering Raymond Chiong The Development of Intelligent Maintenance Decision Support Tools for Palm Oil Mills Ahd. Samad Hasan Basari, Ahmad Fadzli Nizam Abdul Rahman, Siti Azirah Asmai & Zuraida Aba! Abas

318

Fusing Image Features and Information Retrieval Models D. N. F. Awang Iskandar A Comparison of RK-Fourth Orders of Variety of Means on Multilayer Raster CNN Simulation R. Ponalagusamy & S. Senthilkumar

338

323

Performance Evaluation of Face Verification: Different Study Comparative A on Classifiers Shahrin Azuan Nazeer, Mazuki Khalid, Nazaruddin Omar & Mal Kamil Awang

341

Authentication Enhancement for Medical Data Centers Waleed A. Badurik & Shihab A. Hameed

344

315

Computational Models and Systems Two-level Supply Chain with Two-Level of Storage through Information Sharing H. Al. Wee & Jonas C. P. Yu

Human Computer Interaction 326 Head for Robotic Interactive Interaction Amir A. Shafte & Kasyfi M. N.

Saidin,

350

co m er. erg itm

Information Systems

fsp l pd /w ww . p:/

htt o( Vi sit

Zuhana Shaari .

329

332

De m

Aeni Srisuwan, A.smidah Alwi & Nassiriah Chutalip

Social

)

Emerging Technologies and Platform Designing a Web Development Course with Web 2.0 Billy B. L. Lim Requirements Catalogue on Functional Personal Weblog Services

335

X

Success Factors of the Emerging Mobile Advertising Value Chain Chin Chin Wong

356

Data Modelling for Video Databases Mamat, Suriani Affendey, Ali Lilly Hamidah Ibrahim & Fatimah Ahmad

359

Electricity Billing Information System using ICT and GSM Technology H. G. Rodney Tan, C. H. Lee, C.S. Kean & V.H. Mok

362

mo

De isi t

(V

ýýi# h, ,, )

r.c om

litm erg e

fsp

.pd

ww

FULL PAPER

htt p:/ /w

It

f! A%j

Adaptive Threshold for Lane Detection Transform Hough with AZALI SAUDI, JASON TEO and MOHD. HANAFI AHMAD Centre for Artificial Intelligence Computer Science Programme School of Engineering and Information Technology Universiti Malaysia Sabah Locked Bag 2073 88999 Kota Kinabalu MALAYSIA { azali, jtwteo, hanafi } @ums.edu.my

HIJAZI

is essentially the problem of It is a difficult locating problem because of the varying of the lane and conditions that the robot can encounter while moving. surrounding In this paper we present an improvement to our previous lane detection in indoor environment by using adaptive Given a threshold to deal with dynamic environment. real-tune video acquired from a camera mounted on top

This approach, however, is very sensitive to the dynamic of the environment. Thus, in this paper we proposed adaptive threshold to analyze the statistical pattern of the image to be used during conversion of the image to greyscale. This is an improvement to our previous work [I J that relies on static threshold value that tends to fail in a dynamic environment.

of the mobile robot, the gradient of the current lane in detected, so that the near field of view are automatically

2

detection - Lane lane boundaries.

data

er. co m)

erg

(V

threshold,

indoor mobile robot.

1

transform,

mo

Lane detection,

adaptive De

Hough

Keywords:

applied to lane detection is I31. via the Hough Transform Neural networks have been used to attempt to detect lanes on roads not and control vehicles [4], but have difficulties included in their training set. Techniques using tangent vectors have also been shown to be quite robust on wellmarked roads, but can fail when lane markings are not Others have attempted to overcome well defined [5]. lane markings by using multiple problems of differing detectors. For example Gehrig [6] detect hots dots on California highways using a specific detector for hots dots

litm fsp

://w

of the technique to real-time scenario of a building.

isi

the application acquired from indoor

th ttp

from

ww

to the intended the robot wheels are steered appropriately Parameter direction. and space search optimizations tracking strategies are outlined. Results are presented

Related Work

A common technique based on detecting line

.pd

Abstract

Introduction

using matched filters and detect solid lane markings more classical methods.

is a well-researched area of computer vision with applications in autonomous mobile robot and difficult is it However, problem and very a vehicles. remains unsolved. The reason is that it can he very lane difficult markings on various to determine dynamic due to the of the noise particularly conditions, environment, shadows, varieties and inconsistency of the Lane detection

using

Others, such asSouthall et. al [7], proposestochastic methods that able to overcome lighting and road changes while Broggi Is] developed the GOLD system for robust obstacle and lane detection. Earlier, Taylor et. al [91 used a real-time approach for the same problem. Later McCall et. al 1101 examined the use of steerable filters in provide robust lane tracking. Comparative survey has been done in the state of the art of lane detection III].

lane markings. A lane detection system must be able to and filter them to pick out all manner of markings produce a reliable estimate of the robot position and trajectory relative to the lane as well as the parameters of the lane itself such as its curvature and width.

3

Hough Transform

I21 is a technique which, using transforms a set of points some curve representation, defined over the image space to a set of points defined over some parameter space. Points in parameter space image. in instances the the curve of particular represent Therefore, the strategy used by the Hough transform is to instance from of the particular a of points map sets considered line to a single point representing the line in

The

Various lane detection methods have been proposed. They are commonly classified into infrastructure-based Infrastructure-based approaches. and vision-based approaches achieve highly robustness, but construction cost to embed magnetic markers on the road surface is high. In contrast, vision-based approaches with camera on a mobile platform use existing lane markings on the road.

I

Hough

transform

Hough space and, in effect, cause a peak to occur at that point. This approach is considered for fitting a straight

broken line patterns used in central lane markings, or lane markings partially occluded by obstacles such as other robots, people and unexpected objects.

line, as in equation (I) to data.

3.1

(1)

v=mr+c"

Here x and y are observed values, and mnd c represent the parameters. If the values of the parameters are

Algorithm

In the case of a straight line, if there are n points lying on this line, then these points will correspond to a family of lines in the parameter space. All these lines will pass through the point (m, c) in the parameter space. This point gives the parameters of the original line.

the relationship between the coordinates of the point is clearly specified. Rewrite the above equation as c"=ý-mr (? ) given,

In finding the straight line that best fits n points in an image. the mapping from image space to the parameter space is used. In this approach, the parameter space is represented as an array of accumulators, representing discrete parameter values. Each point in the image votes for several parameters. according to the transformation To find parameters that characterize the line, equation. one should detect peaks in the parameter space. These steps are summarized as below:

and, assume that m and c are variables of interest, and x and y as constants. Equation (2) above represents a straight line in the m-c space. The slope and intercept of this line are determined by x and y. A point (x, v) corresponds to a straight line in mc space. as shown in Figure 1. In practice, the polar form of the line (equation 3) is used rather than the explicit form to avoid problems with lines that are nearly vertical.

Quantize the parameter space appropriately. Assume that each cell in the parameter space is an accumulator. Initialize all cells to zero. For each point (x, y) in the image space, increment by I each of the accumulators that satisfy the equation.

" " "

(3) p=xCos0+vsin0 If there are n points lying on the straight line, then these points will correspond to a family of straight lines in the parameter space, as shown in Figure I. All these lines will

Maxima

in the accumulator array correspond the parameters of the model instances.

to

m)

"

litm erg

er.

co

pass through the point (m, c) in the parameter space. This point gives the parameters of the original straight line,

.pd ww

in

Lane Detection System

fsp

4

The overall ttp

://w

y

Hough Transform

implementation

steps of the system are shown

De

mo

(V i

sit h

in Figure 2 ti)r the purpose of testing the system. The video inputs are taken from a forward looking camera, on top of the mobile robot. Step 1: Image acquired from the camera.

ý /

Step 2: Image is threshold to be converted into binary form.

0Ö` Figure I. Image-to-parameter

space mapping of a point in the Hough transfomi

Step 3: Edge detection is applied to find points of candidate line.

In digital

image processing, equation(3) can be interpreted the total pixel intensity as evaluating along a line parametcrised by p and 0. Other interpretations of this

Step 4: Perform edge thinning process to reduce edge pixels in the image.

transform view the process as that of accumulating votes for each possible line in the image. Lines for which a high number of votes accumulate result in occurrence of peaks in Hough space.

Step 5: Hough transform is applied to find the straight line that best fits the candidate line.

One of the main advantages of the Hough transform is its robustness to noise and occlusion. This is due to the fact that each image point is considered independently of the others. hence. within reason, the removal of valid edge

Step 6:

Find maxima, remove fake lines, and pick the best possible line, then send signal to steer the robot to the intended direction.

or the introduction of' spurious noise points only in Hough alters peak intensities space. A direct consequence of this is the techniques capability to detect points,

Figure 2. Steps in lane detection

2

These inputs are then fed into the lane detection system to determine the state of the robot and its path. A Sobel edge detector is applied on the image to find points of candidate line. The greyscale colours in the image are reduced to black and white using the adaptive threshold details before operator so as to remove unwanted

the method via Hough Transforms is tested using real-time data acquired on camera mounted on top the mobile of the robot. The scenario is an indoor environment Technology, i. e. School of Engineering and Information the corridor with existing permanent lane marking on the

it thinning operation.

applying

floor. The

threshold

value a statistical

performing intensity

values.

Here

is

by selected on the sampled pixel

dynamically

analysis the

histogram

Results

Experimental

5

occurrence

of

shadow

and non-uniform

lighting

condition. Below are images taken from these tests that demonstrate the system working under these conditions. Figure 3a is the original image acquired from the camera. This image

and cumulative

of the gradient magnitude image can he created. This approach, the top IO`%%to 20'%% of the largest gradient values are selected as edge points. histogram

is then threshold Figure 3h.

to get its greyscale

values as shown in

Because of the nature of the Sobel operator, the thinning operator must he applied to reduce the lines with several pixels width to a single pixel width. The key method is the I lough transform which involves the conversion of pixels in the image from in age coordinates (x, y) to parameter space (p, 0), and then search in the Hough space to find the long straight lines, which are lane marking candidates. Finally, the best fit candidate line will trigger signal that steer the robot to the intended direction. fast detection, some more post order to perform processing on the filter results is performed. First, only the filter candidates within the vicinity of the lanes are used in updating the lanes. This removes outliers from Moreover, other objects or obstacles in the environment.

om er. c erg litm fsp

.pd

thus speed up

Figure 3a. The original

ww

process reduce computation,

://w

this filtering

)

In

mo

(V

isi

th

ttp

the calculation.

picture acquired from the camera

for each lane, the first and second moments of the point candidates are computed. Straight lane markings De

Secondly,

should he aligned so that there is it high variance lane heading direction and a low variance in the direction. Outliers are then removed based on Thirdly, only lanes that start from the statistics. part of' the image are considered, fake lanes.

Finally, cluster

lake

lines

are removed.

need to be grouped

in the other these lower

to reduce number

At

this

of

stage, the line

as one lane marking

and other

lines are lines need to he deleted. The candidate in the image from to their position sorted according left to right. Then for each line group consisting 01' closing fake

lines, the most possible line as the lane marking is distance lines between two fake (the other selected and lines and their count numbers are used as criteria to judge

straight

fake lane is line this marking) a not whether or

is deleted.

because the alCOrltllll uses a local search about the lanes In testing, it was for candidates. it requires initialization. the lane tracker position to initialize and sufficient trajectory lane).

to zero

(corresponding

to the centre

Figure 3h - Grayscale of' image 3a In Figure 3c. Sobel operator is chosen for edge detection. however, This edge detector, lines. thick produce which will produce many fake lines later in the postprocessing stage. Therefore, edge thinning process is done to reduce number of edge pixels for better computational speed. Figure 3d shows the output of edge thinning

of the

process.

3

\

\ ý \

Figure 3c - Output image of Sobel edge detector

Figure 3f -Final result after fake lines are removed The performance shows that the method is robust to varying condition of the environment. However, much improvement can be done to speed up the processing, particularly in the pre-processing stage and post-filtering. One

particular challenge in indoor environment is that the lane marking are not uniform in different areas of the corridor.

6 \

Discussion

fsp

The output of edge thinning -

objects or obstacles.

process

://w ww .pd

Figure 3d

litm

erg

er. c

om

)

Lane detection is often complicated by varying markings in indoor environment, clutter from other objects and from other changes, and occlusion shadows, lighting

in indoor environment the problem of lane detection is presented. The system demonstrates that it is capable of handling varying conditions of the environment. Future work will emphasize on improving the technique to speed better to up computation and algorithm reduce

th

step, the candidate lines are sorted and grouped together. Only one best possible line from a group of lines is

complexity.

computational

This

process deletes many fake lines, and produces only a handful of thin straight lines as shown in Figure 3f. As the lane detection system traverse each line from left to right, it will pick the middle as the most possible line. chosen.

to

(V isi mo

De

the Hough transform produces the candidate lines as shown in Figure 3c. The initial output. however, contains many fake lines. Thus, in the post-processing

work

ttp

Finally,

In this paper, a preliminary

data

to

parallel

a

more

computation

Wireless

powerful for

transmissions

processing

faster performance

power

of' and

will also he

considered.

References III Azali Saudi. Hough Detection "Lane using Transform for Indoor Mobile Platform". Proc. of the 3rd Int. Conf. on Artificial Intelligence and Engineering Technology, 22-24 Nov 2(X)6.

121 K. Duda and P. Hart. "Use of the Hough transform to detect lines and curves in pictures". Communications of the ACM, 15: 11-15. 131

J.

McDonald,

"Detecting tracking and road in Proc. of the markings using the Hough transform", Irish Machine Vision and Image Processing Conference, 2(x)I, pp. 1-9.

Figure 3e - The initial output of Hough transform produces many fake lines

141 I). Pornerleau. Ralph: "Rapidly adapting lateral position handler." Proc. IF.F.F. Symposium on Intelligent Vehicles. September 25-26.1995.

4

[5[ H. Furusho, R Shirato, M. Shimakage. "A Lane Recognition Method Using the Tangent Vectors of White Lane Markers, " 6th International Symposium on Advanced Vehicle Control, Sept. 9-13,2002. [6] Stefan Gehrig, Axel Gern, Stefan Heinrich and Bernd Woltermann, Lane recognition on poorly California", in Bot Dot The problem roads structured Proc. of the 5th Conference on Intelligent Transportation Systems, 2002. 171 J. B. Southall and C. J. Taylor "Stochastic road Computer Conference International on shape estimation" Vision, pp. 205-212, June 2001. [81 M. Bertozzi and A. Broggi. "GOLD: a Parallel Real-Time Stereo Vision System for Generic Obstacle Image Transactions IEEE " Detection. Lane on and Processing, 1997. [91 C. J. Taylor and Jitendra Malik and Joseph Weber. "A Real-Time Approach to Stereopsis and LaneFinding. " Intelligent Vehicles 1996, pp 207-212.

litm erg fsp pd w. /w w

De

mo (

Vi sit

htt p:/

[I I1J. Kosecka, R. Blasi, C. Taylor and J. Malik. "A Comparative Study of Vision-Based Lateral Control Strategies for Autonomous Highway Driving". In IEEE Int. Conf. on Robotics and Automation, pp 1903-1908, May 1998.

er.

co

m)

[101 J. C. McCall and M. M. Trivedi, "An Integrated, Robust Approach to Lane Marking Detection and Lane Tracking", in Proc. of IEEE Intelligent Vehicles Symposium, Parma, Italy, June 14-17,2004, pp. 533-537.

5

Benefits of In-House e-Procurement System NORHIDAYAH MOHAMAD and WAN KHAIRUZZAMAN WAN ISMAIL Institute of Technology Management and Entrepreneurship Universiti Teknikal Malaysia Melaka, 75450 Ayer Keroh, Melaka Faculty of Management & HRD Universiti Teknologi Malaysia 81310 Skudai, Johor MALAYSIA [email protected] edu.my, [email protected] my transformation due to the dynamic developments of information technology and telecommunication (ICT). The advent in the Internet has enable procurement to be made online and enhance the effectiveness and of doing business. E-Procurement system efficiency gives a new platform for companies to run their business via online locally and globally [3].

Abstract in electronic procurement - The performance (e-Procurement) the may affect efficiency and effectiveness of the procurement process of an However, organisations organisation. were reluctant to implement this system due to its intangible advantages benefits and the difficulties in using and potential Thus, this research measures the of an in-house e-Procurement in e-Procurement system and identifies the differences based on demographic factors. Multiperformance method approach was used where distribution of

software. level performance

co

er.

litm erg

fsp

.pd ww /w

htt p:/

interviews ss"ere questionnaires, and observations conducted accordingly over time. It was found that the level of performances of the system was moderate at between 50.8%. The findings showed a difference performance exists, education level and e-Procurement significant at 0.05. Few harriers were identified but they low (2.31) and were overcome were considerably during the early stages of'the system's development The

De

mo (

Vi

sit

system is relatively cheaper and perhaps provides more advantages than those offered by the existing system in the market.

Keywords: e-Procurement, information Internet, e-Commerce.

I

The results from a survey among companies in Penang showed that there was strong interest in the development and potential of e-Procurement system. By and using standardized protocols, efficiency effectiveness may be achieved and will enable companies Most to participate in the global supply chain [4]. companies were positive in terms of their acceptance towards using new e-Procurement technology as a The benefits provided by ecompetitive strategy. for Procurement system were the main driver The system companies to implement this system. allows suppliers to introduce and promote their new products through web pages, to receive, manage and process purchasing orders and receive payments. Furthermore, companies can publish their product catalogues innovatively and furnish quotations, tender documents and submit tenders online. m)

available

technology,

2

Introduction

Literature review

There were some research studies on the level of Internet used in this country. From the research on all the states in Malaysia, the average usage of Internet is just around 25 percent and Malaysia still far behind if compared to the developed countries. As the country increases itsl('Tuptake, some of the local companies in this country suffered huge setbacks simply because they were not ready to face the ongoing challenges. Forexaniple. the Malaysian Small and Medium Industries (SMI) Association revealed that only 30 percent of the local SMIs have their own websites, and even worse, most of these sites were not regularly updated. The implication here is that the companies still lag far behind in the acceptanceof this new but necessarytechnology ]5].

Electronic procurement or e-Procurement is one of the information technologies that apply Intemet as a medium of transaction. The e-Procurement system enables suppliers to sell their goods and services online to their customers. It also engages the suppliers and customers directly in a global commerce environment. The use of Internet in Malaysia saw an increasing trend and the number of broadband Internet users have surged to 1.3 million residents in 2006. This phenomenon reflects the importance of Internet not only to individual users but However, Malaysia is still also the industries [I]. lagging behind in Internet-related infrastructure when compared to the western countries. The government has been working hard to ensure that it is not left behind [2] and continues building the knowledge economy in order Traditional to uplift the country's competitiveness. Malaysian retail industries such as shopping malls, financial and security sectors have been undergoing rapid

F.-Procurement system is one of the technologies that assists companies in managing purchasing process. the acceptance of the system in Malaysia is still low and yet to capture the interest of the masses 121. The intangible

6

Table 1. Factorial analysis and Eigen values for eProcurement benefits

advantages among the suppliers were one of the reasons why suppliers were hesitant to use this system [6]. Understanding of the e-Procurement system is imperative to ensure the ability of each company to compete in future. Another problem faced by organisations in implementing a procurement system was the difficulties to use the existing software in the market [7]. This was due to the difficulties in integrating the software with their existing procurement process and the longer time needed to design such a system.

e-Procurement

ctors

Factor 3

Factor I

Factors*

Bene fits for user

Reducetransactiontime

0.828

0.173

2.279E02

Deliver information to internal

0.716

0.142

0.352

Reduce transaction costs

0.708

0.403

-1.44E03

Increase products/services

quality to internalcustomer Increaseflexibility to internal customer'schangingneeds

0.699

0.199

0.308

0 695 .

-820E 03

0 330 .

Reduce order processing

0.694

0.345

errors E-Procurementsystemwas sufficient

-9.39E-03

0 664 .

0 140 .

0 187 .

0.661

0.553

0.104

0 644 .

0 557 .

0 149 .

0.617

0.465

0.274

customeron time

Most researchon e-Procurement in Malaysia focused on the application of the technology in the industrial sector. However, research on the e-Procurement system which is developed in-house was insufficient. As has been mentioned, the purpose of this research is to measure the level of performance for an e-Procurement system in-house by Telekom Research and developed Development Sdn. Bhd (TMR&D) to cater its own procurement needs.

Improve overall service quality to internal customers Increase communication with

internalcustomers Meeting internal customer expectations

Methodology

qualities

0.600

6.329E

0.297

-02

Increasereliability of informationto internal

0.598

0.344

0300

customers Reduce paper work

0.511

0.407

0.144

.pd fsp

litm

erg er. c

om

)

Benefitswith reliard to the relationship suppliers 0.150 0.872 8.199E-02 Improvecommunications Improve partnership 0.242 0.152 0.868 0.403 0164 0.703 Improveddatasharin

mo (V i

sit h

ttp ://w

The data for the study was collected through questionnaires, interviews and personal observations. One of the authors was given the permission by the case study organisation to stay for almost half a year to conduct this study. The survey instrument was adapted from Quesada [8] and has been tested for its validity. Semi-structured interview was selected because it gives more insights particularly on the problems and barriers of the system. Participant observation was carried out to understand the research phenomena more in-depth and has allowed researcher to gain hands on experience in using the e-Procurement system.

Increasematerialsservice

ww

3

Lo=i

benefits

Benefits to the organization 0.191 Reduce inventories 0.131 8503E Reduce maverick buying 0.209

De

Reduce number of suppliers Ei en Value % Variance

Data was collected from secondary sources via company's documents, journal, articles, seminar papers, newspaper cuttings and also past thesis. Descriptive analysis including means, standard deviation, multiple regressions, ANOVA and t-test were used. For qualitative analysis, data was filtered and all unnecessary data was deducing data, displaying in the of process eliminated data and conclusion and verification of the study [9].

6.267E02 9.718 44.172

2.186E -03 2.103 9.560

0.828 0.812 0.802 1.613 7.332

Factor I: Benefits to the users Factor 2: Benefits with regard to the relationship with suppliers Factor 3: Benefits to the organization

Factor analysis was carried out at the early stages of the research. Table I elaborates the results of factor in factor Eigen each and the acquired values analysis and loading factor for each item in the probability for the benefits of e-Procurement.

benefits for users of the system included reduction in the transaction time (0.828), cost (0.708), paper work (0.511) and decreasein order processing errors (0.649). This system was able to increase material service qualities (0.600), services (0.661) and reliability of information to the internal user (0.598). From the weight value, reduce transaction time was the most significant for e-Procurement system as at TMR&D.

From the factorial analysis on e-Procurement benefits, the benefits were grouped into three factors: benefits to user (9.718%), benefits to suppliers (2.103%), and benefits to organization (1.613%). These three factors describe 61.065% from the total of variance for eProcurement benefits. All the variables have loading values of more than 0.40 in each factor.

E-Procurement system is also beneficial to the case company in reducing (i) inventory (0.828), (ii) maverick buying (0.812), (iii) number of suppliers (0.82), (iv) delay in the delivery of products (0.465) and (v) cost of materials services (0.402). When compared to all the loading factors higher than 0.8, reduction in the inventory, maverick buying and number of suppliers were the three

The

4

Analysis and findings

7

main factors that contribute to the benefits received by the organisation.

Table 3. Statistical value for e-Procurement benefits based on demographic factors e-Procurement

Statistical

benefits

4.1 Level of e-Procurement

Demographic factors 19cnctits to

Table 2 shows that the benefits of TMR&D eprocurement system is at a moderate level (50.8%). Many researchers agreed that e-Procurement system provides numerous benefits to its user (e.g. [10], [I I]). In this research, the benefits were divided into three categories: benefits to the users, benefits with suppliers and benefits to the organisation. The comparison of frequencies among these three benefits indicates that the benefits to the users was the highest level at 42 followed by the benefits to the organisation and benefits with regard to the relationship with suppliers with frequencies of 46 and 47 respectively.

or

mmbon

Benceits to SU

ICTS

licncl its to

(: ender

value (s1gIniOcance)

Education

Experience

Designation

0 485

0 734

0 179

0880

0911

0 228

0 672

0 306

0 765

0 042

0 930

0313

Internalusers

4.2 Effects of e-Procurement

system

Results from the interviews with the management indicated that huge benefits can be earned from the implementation of e-Procurement system. For TMR&D organization, the benefits from the system were:

Table 2. The level of performance of e-Procurement Levey Benefits Renefits to organization

Hi h

Moderate

IoM

13

46

30

I. 2. 3.

Benefits to user

42

41

S

Henetits from suppliers

33

47

11

88

134

Total

43

4.1.1 Performance demographic

of

e-Procurement

erg er. co m fsp litm .pd

5. 6.

ww p:/ /w

De

mo (

Vi

sit

htt

The head of TMR&D procurement unit iterated that this system was introduced to facilitate the procurement process and at the same time provide advantages to the users. With this system, the users felt that executing purchasing was easier as compared to their previous experience with the traditional procurement process.

)

4.

The procurement process was more effective and efficient Increased profit c-Procurement was a new technology and in tandem with the current needs for organizational competitiveness More transparent and easier to monitor the procurement process Reduced the abuse of power Better decision on selection of the best product and services

and

factors

Table 3 shows the significant value for e-Procurement performance based on demographic factors. There was no difference between significant c-Procurement performance and gender, designation and working experience factors. This may be because this system was designed for all the "I'MR&D workers and more userfriendly. Studies have also found that there were no differences within behaviour and gender (121 and that gender and designation have no effect on the purchasing aspect 1131.

All these benefits were the motivator for the company to develop and adopt its e-Procurement system. From all the listed benefits, the main benefits were transparency, less purchasing process time: reduced purchasing time, and also lower cost and better product specifications.

4.2.1 Transparency Transparency is an important aspect in running a business especially in purchasing. Attention on transparency can make the purchasing process runs smoother. From the interview conducted, transparency was crucial for the managers to ensure that there is no misuse of power. Conducting purchasing online may increase the level of control of the system which is more by he it to may accessed avoid suspicions since and open any of the south.

On the other hand, there was a difference between eProcurement performance (benefits for user) and education level. Post floc test indicated that this difference exists among Masters and Bachelor degree holders. Leticia (2002) argued that the higher the education level, the higher the preference toward using the Internet or Therefore, the authors believe computer [14,15,16J. that the workers with higher education felt more comfortable using the c-Procurement system because education level is related to the acceptance of new technologies [ 19J.

By having this e-Procurement system, all inf rmation about purchasing such as the list of suppliers and other purchasing activities are kept in the system database. In addition, the system is also capable of producing latest reports on suppliers' performance and updating of Thus, every activity related to purchasing activities. procurement can be monitored by all the l'MR&I) staff while getting crucial data about the company's suppliers.

8

4.2.2

Shorter procurement

By developing its own system, the procurement unit is able to make changes immediately according to the management needs. These changes and upgrading are done without maintenance cost and is therefore preferable by the management in the short span period of time compared to the usage of the existing software in the market.

cycle time

has also system of e-Procurement adoption for time the purchasing activities. process shortened Report from TMR&D procurement unit shows that the time taken for a transaction process has been reduced from 3 weeks to 3 days. The

The short span of time has indirectly helped the to obtain their research researchers at TMR&D time faster the same at while equipment materials and help produce their research products according schedule. Reducing the procurement cycle time lessens the idle in delays for a can avoid researchers products and period particular R&D project.

Conclusion

5

ICT plays a significant role to enhance a company's effectiveness and efficiencies and enable companies to compete in the global market. The use of ICT applications such as Internet in procurement is one of the strategies to upgrade and improve procurement process. Previous studies have already proven that the use of ICT in eProcurement provides many advantages for the improves E-Procurement the procurement. company's purchasing activity and makes it simpler than the traditional method which is more complex, cumbersome and inefficient.

Previously at TMR&D, the traditional procurement this is and to purchasing make used method least be three months to at applied purchasing needs Normally, before any R&D project commences. purchasing for R&D activities may reach millions of ringgit. The amount includes purchases of machine and On foreign the local from other countries. and equipment hand, careful evaluation must be conducted to ensure that the products purchased fulfill the specification to support the The that shows analysis the research. and conduct system is capable to increase the level of efficiency of the company's procurement.

fsp

References

De

mo

(V isi t

htt

p:/

Reduction in the overall cost is one of the reasons why In system. c-Procurement adopt majority of organizations the case of TMR&D, the cost is further reduced because its e-procurement system was developed in-house. In for it the cost the reduces system, terms of the usage of for information through as as well mail sending the Users are only reports. proposal printing and preparing forms and the the online to complete required he the to company's advertised will applications it cheaper, simpler and making registered vendors, into cost and time savings. translated eventually

litm erg

er.

cost

w. pd

Reduce procurement

/w w

4.2.3

co

m)

Overall, internally developed e-Procurement serves as an alternative to those companies that are eager to improve their purchasing process but are constrained in terms of capital and expertise needed. Although there were few weaknesses in the e-Procurement system, it should be a hindrance for companies to implement the system as it proves to offer more advantagesand benefits.

4.3 Advantages Procurement

of developing

[II Abdul lah, A. B. Opening remarks of'a meeting of she for Information Technology, Council National http: //www. pmo. gov. my/ W ebNotesA pp/PMMain. nsf/0/6d6 Access 3e 18205df5b948257I 5d000f7aeb? Open Document date: 27 June 2006,2006.

Ramayah, T. and Muhamad, J., "Intention to The purchase through the world wide web Malaysian experience," The Third International Conference on Electronic Commerce Engineering (ICe('E2003), Hangzhou, China, 2003. [2]

in house e-

system

Hussain, A., Mohamad Kamal, F. and Afzan Musa, [3] in The R., to suppliers study of e-Procurement system internally enables Developing e-Procurement system International Kedah Proceedings Penang of and area, for its the system. own maintenance organisations to conduct 2005. E-Commerce, Conference on Different forms of maintenance, updating and upgrading by be the out staff carried can which the system of frontiers A., The [4] Helmi, e-Procurement: of from for technical outside. having assistance to wait without RosettaNet Malaysia, Proceedings of International 2005. E-Commerce, Conference in on important making The maintenance aspect was also decision whether or not to implement a certain system. This SME C. S., T., Lim Y., Mohamed, [51 Ramayah, in eand in technology changes require is because changes for in Implications Malaysia: Readiness planning and Therefore the technology. the new to suit organisations Vol. 11, implementation, Journal Management, Sasin of that they continuously system users must always ensure 1, 103-120,2005. No. different pp. The types to of needs use fulfil the current needs. depending from to differ another company one system may C. Segev, A., "Impact Gebauer, J., Beam, [61 the of and top management. on the choice of Internet on procurement," Acquisition Review Quarterly

9

Vol. 14, No. 1, pp. 167-181,1998. [7]

Davila, A., Gupta, M., and Palmer, R. J., "Moving procurement systems to the Internet: The adoption and use ["111"o/van technology of e-Procurement models, " fanaf ement Journal, Vol. 21, No. 1, pp. 11-23,2002. .1 [8]

Quesada, G.,

meni iechnoh>gies. .l .stach' oje-l'rrxure procurement practice. procurement performance and their relationships, PhD thesis, University of Toledo, 2004.

[91

Miles, M. B. and Iluberman, A. M.. An expanded sourcebook-qualitative data analysis, 2"' cd, London, Sage Publications. 1994. 1101 Frye. D. W.. li-l'ro uremrnl in Ihr l'riraIe and Public S'ectorc,PhD thesis, George Mason t niversity. 2004. [111

('room,

S. and Johnston, R., "E-service: Enhancing internal through customer service e-Procurement, " International Journal of k'rrice lndnshl' tlanagemenl, Vol. . , . 14, No. 5, pp. 539-555,2003. 112]

Gattiker, U. and Illavka, A., -Computer attitudes and learning performance: Issues fimr management education and training, " Journal of (hgani: ational Behavior, Vol. 13. pp. 89-101, I )K)2. P.('.,

In empirical .

inresli}; utirm

inh, the

)

Sharon,

co m

1131

1151

Rugimbana,

116]

Robich,

1171

I3akri,

(V i

sit h

computer and the Interne:. Forum, at fact sheet. html. Access date: De mo

114 Leticia, V., Latinos, Latino Issues http: '/www. lit.. orl,, technology 22 June 2006,2_002.

ttp ://w

ww

.pd

fsp l

itm erg

er.

henreen electronic rrnnmrrre rt'luriunshih and stillt' and local gurernntenls'. cule.c and use nn'e.c, NO thesis. College of Business and F:conomy, l hiiversityof Kentucky, ? (X) I.

R,.

"Predicting teller automated machine usage: the relative importance of perceptual and demographic factors, " lnkrrnurional Journal HunA of . t larAt'nng. Vol. 13, No. 4, pp. 26-32,1995. /k'rc'c'J)IiO is .1 vfudi 0J c -hunkin, L securim School i.c.cues. PhD thesis. of rind i u.clumcr cnlis/action Business, l'niversity, Argos) _2005.

l1-mhu, lamun leknuluKi (li A,dui{yan /n-Ae, ju .crAlur li(luslri leAm, lrgi rin}; Xi , li 1/aLnsia. 11h1) Malaýsia, 2003. thesis, Universiti leknuluai ti.,

10

HCl for Non-Literates: An Experience from Sambad Project SAGUN

GANESH B. GHIMIRE, PATRICK A. V. HALL, DHAKHWA, PRAKASH MANANDHAR and ISHWOR THAPA Sambad Project Madan Puraskar Pustakalaya Patan Dhoka, Lalitpur NEPAL

[email protected] org applications focus mainly on literate users preventing the use of computers by the illiterate. So we surveyed, interviewed and conducted focus group discussions in different parts of the country to gather information on what type of applications could be useful to the non-literate people. From our survey that a strong thirst for knowledge and rich indigenous knowledge was evident in the rural communities. It was obvious that this information is needed to be delivered text free, with pictures and speech, and with a user interface that is easy to use. Hence we decided to build a Multimodal Web Editor and Browser which uses `speech as data' and has support for touch screens. Available

Abstract - Trends in social computing has shown that Internet users have evolved from consumers into producers and contributors. However, these trends are literacy level Nepal, like fir the where true countries not is just above 50%. Literacy has been a prerequisite for the use of computers. Still, the illiterate societies have a hold knowledge thirst and valuable indigenous of strong knowledge that can he shared on the Internet, but there is no proper Human Computer Interface for them. Software which allows illiterates to create and browse their multimodal webpage using speech and images can he a solution f rr this problem, if there is a better design. Sambad attempts to build an interaction design which interaction exploits visual, aural, observational and tactile learning human behavior for access of computers fir non-literate people.

) er. co m

litm erg

fsp

o(

Introduction

software working text, providing all provided for text, public domain and

Speech needs to be compressed and decompressed, and many of the compressions algorithms like MP3 are subject to patent and license fees. We want to be able to do standard editing operations on speech of which searches and cut and paste seems to be the most challenging. In section 2 we discuss the various choices we had to make, and the architectural framework that we used to guide those choices.

De m

I

Question 1: Could we put together with speech and images but no written the functions that would have been taking standards and modules from the open space?

.pd ww ttp ://w

Vi

sit h

Human ICT for Rural Communities, Keywords: Computer Interaction for non-literates, Multi modal web authoring and browsing tool.

We have been addressing two fundamental research questions:

New IT enabled devices have been the reason behind the emergence of new types of social organisation and Online developed in the world. networks, cultures communities, blogs, pod casts are some of the new examples of these cultures. literacy seems to be a prerequisite for for interfaces because software are user computing is 48.6% [11, in Nepal Literacy based. text conventionally thus, half the population is denied access to, and benefit from computers, before other issues like connectivity, into We taken use the account. are on so economics and interchangeably, and 'non-literate' terms 'illiterate' though we prefer the latter because it is less pejorative. fhe sense we intend is being unable to read and write literacy as a broader the views of text, not and written being able to understand pictures, use technology, and so on. However

Question

2:

Could

functions we present these addressed in question 1 to non-literate users so that they could understand what to do and do it effectively? Illiterate users will also be technically inexperienced and never have seen a computer before, though they may have used televisions and mobile phones. This meant that we had to be open to different kinds of interfaces, questioning ourselves continually about what would be an appropriate interface, evaluating and modifying our interfaces in cycles of user-centered design. We give a flavour of those interfaces and how we evaluated them in section 3, though a fuller description is given in Sambad design of computer interfaces for non- I iterates. [21 the -

We want to help illiterate people to. join the information did they that that needed software recognise society, and instead but 'text-free' depend was which writing upon not interacted with the user though information and stored speech and images.

Neither of these questions are trivial questions.

11

The System

2 2.1

" ýd., ýIo,,,oý.ý -7 f

Software System Overview

Sambad is a text free system, Multimodal Web Authoring designed to provide all and Browsing Tool, the functions that would have been provided fir text. It can be broadly divided into five parts: file management, UI

Custom

synchronizer initialization component

Widgets,

file importer, as login, which

No 1

Is user valid

file

editor and as well conies under the I ). File Playersubprocess (see Figure is used commonly by file editor, file and Custom UI Widgets. Sambad reuses

Village

Read User Input

nianagenient various open source projects and open standards for lower level functionalities like SPEEX codec [3], XML, speed adjustment etc., and uses different patterns to integrate these components.

audio

Metapher

I I Cupboard Metaphor

design Synchronize with Server

StNfPTIUnNQ"

Synchronie

wdh

I

server

I

ý

ý

ý

N.,

Synduoniic with server

4

litm

erg

er. c

om )

ý

Vea

Diagram of the System

th

(V isi

diagram in Figure 2 shows the basic The activity in business and operational of components workflows Sambad. When the program is started, after the successful login Sambad client synchronizes with remote Server. Synchronization is the process of downloading new files

Figure 2. Activity

ttp ://w ww

.pd

fsp

Figure I. Block Diagram

De

mo

Manage File i

from

the server and uploading newt created or edited files to the server. Users can start doing various activities as shown in the use case diagrams that folloss.

Import File

Synchronize

2.1.1

File Management Edit File

file system. kith Sambad file system is a hierarchical file the conventional three levels which simplifies hierarchy. This hierarchy is maintained using an XMI, file. The village is a root under which there are ditiarent places (e. g.: home, school, etc, ). I hese places have cupboards. A user is provided ssith and public cupboard in one's home directory, 141 presentation. record where one can create SMII. tiles and export them. There will be speech, import domain experts to administer places apart from home. private private

Play File

and public

The system currently supports 4 types of files namely SPEEX, image. SMII., and album. SPEE:X files are the audio files encoded in SPFI: X format that the system Sambad supports all file uses to play instructions. formats for image files that lava Sssing supports. Album . is a collection of'these image tiles.

I igurr 2.1.2

Usc Case I)iagram

ul' 111ctiN sICnt

I'Iav File identities

all file presiessing as I ilc flay. Rased on the life types. there are 4 dii erent Ispes of players nanrels image %ie\+er, album \iewwcr. SI'I: I`.X player, and SMII. placer. Image and album vies%er are I'NG. (ill-. that display images in tl'I'G, components . RMI', etc. format. SI'I`. I`.X player is an audio pla\er format and has the \%hich decodes audio tiles in SI'I'TX Santbad

12