Research

MPhil/PhD Research

HK Salinda Premadasa

Start Date: 01/10/2009

Completion Date: 24/03/2014

Email: salinda(at)ccs.sab.ac.lk, salinda.premadasa(at)gmail.com

PhD Research Title: Demand-based Mobile Learning Network

This research focuses on the open-source and secure mobile learning environment with short messaging system which facilitate two-way communication with data concealment and extensive message compression mechanism with cryptographic protection that can be easily integrated into an LMS to provide an interactive learning experience to the user community.

Authenticated students can perform learning activities such as content downloading, group discussions, online quizzes and assignments in the mobile learning environment using their mobile devices. Also, authenticated teachers can create text messages using a mobile device connected to the Internet. The text message can insert into the LMS integrated database and SMS Daemon periodically checks and sends it out through the SMS gateway to the recipients. If the message is extensive than 160 characters, it is compressed and given cryptographic protection by the system prior to inserting into the database. The recipients who handle mobile devices decompress such compressed message into an original form of texts to view them. The UDH associated port addressing technique with short messaging service is used to conceal and exchange data relevant to a particular course unit concealed in the message body itself. Subsequently, software installed in the user’s mobile device captures this message and sends back the reply message to the appropriate course unit allowing both teachers and students to view messages sent and replies received pertaining to a particular course unit.

Finally, the research proved that the students’ preference in studying with a group while using mobile technologies for their academic activities, system relevance and interoperability of the two-way text messaging with data concealment and educational impact on improving the learning environment. Also, the message compression ratios, compression time variations for the various types of standard file sizes and message delivery time with the comparison of standard SMS, illustrate that the relevance of using compression mechanism to mitigating major drawbacks such as cost per message in the mobile learning environment.

Publications:

International Refereed Journals

HKS Premadasa and RGN Meegama, “Mobile Learning Environment with Short Messaging Service: Application to a Campus Environment in a Developing Country,” Campus Wide Information Systems, Emerald Publication, vol. 30, no. 2, pp. 106-123, March 2013.

National and International Conferences

HKS Premadasa and RGN Meegama, “Extensive Text Message Compression with Security on Android Mobiles,” in IEEE International Conference on Advances in ICT for Emerging Regions (ICTer), Colombo, Sri Lanka, 2013, pp. 80-83.
HKS Premadasa and RGN Meegama, “Implementation of Concatenated Short Messaging System in a Campus Environment,” in IEEE 8th International Conference on Computer Science & Education (ICCSE), Colombo, Sri Lanka, 2013, pp. 207-212.
HKS Premadasa and RGN Meegama, “Transmission of Concealed Data in Short Messaging System,” in National Conference on Technology & Management, Sri Lanka Institute of Information Technology (SLIIT), Colombo, Sri Lanka, 2013, pp. 120-124.
HKS Premadasa and RGN Meegama, “Concealed Data in Two Way Text Messaging: A Framework for Interactive Mobile Learning Environment,” in IEEE International Conference on Advances in ICT for Emerging Regions (ICTer) , Colombo, Sri lanka, 2012, pp. 122-127.
HKS Premadasa and RGN Meegama, “Mobile Learning Environment with Short Messaging System,” in National Conference on Technology & Management, Sri Lanka Institute of Information Technology (SLIIT), Colombo, Sri Lanka, 2012, pp. 95-99.

TDG Geethika

Start Date: 31/03/2010

Email: gilmini(at)dcs.ruh.ac.lk

Supervisor(s): Dr. Prasad Jayaweera

MPhil Research Title: Development of a Unified Framework for Business Motivation Driven Business/IT Alignment – An application in Healthcare service sector

Enterprise modeling can be seen as a complex and time consuming process involving variety of participants. One promising approach to tackle this situation could be the separation of concern during enterprise solution designing. As a result it may be needed to develop different models during the different phases of the development workflow such as Goal models, Value Models and Business Process Models. Since all these methods are describing each layer discretely, it is often hard to find relationships between different modeling concepts at different enterprise modeling layers such as the ones listed above. Therefore, there is a greater chance for miss-interpreting or miss-communicating of captured semantics between aforementioned layers (Goal Layer, Business Value Layer and Business Process Layer). As a result miss-alignment could cause bad/inappropriate enterprise solution designs that may not be able to meet intended business requirements and/or also to sustain in ever changing business environments.

Although the same consequences experiences in many e-Solutions used in the healthcare domain, however still there is a lack of systematic guidance to provide alignment between aforesaid modeling aspects. Also Service designing is most preferred in healthcare systems because the patient receives the service as a bundle of services which are provided by multiple stakeholders. Thus it is necessary to find how the services offer through shared environment to maintain the interoperable behavior of the services while ensuring higher level motivational requirements of the different service providers.

More specifically there is a need to study existing enterprise modeling ontology that facilitate different modeling layers and then to formulate precise conceptual definitions for selected artifacts. Further the fundamental requirement in achieving alignment is to understand and define existing relationships between different artifacts at different layers.

The significant issue address by this research is how to address the alignment between lower level technical requirements and strategic requirements of the business in system designing that is providing systematic guidance to business IT integration. Also the work needs to be applied to adopt standard existing framework of healthcare domain towards systematic e-service identification with addressing interoperable behavior of it.

Publications:

Conferences

Dantanarayana G., Jayaweera P., Value Oriented Service Designing for Successful e-Health, in Proceedings of the 9th International Conference for Internet Technology and Secured Transactions (ICITST-2014), (2014, December)
Dantanarayana G., Wickramage C and Jayaweera P., Goal Oriented Value Object Classification for Healthcare Service Development, in Proceedings of the Fifteenth International Conference on Informatics and Semiotics in Organisations (ICISO 2014), (2014 May)
Dantanarayana G., Wickramage C and Jayaweera P., Framing Services based on Value Activities in Healthcare, in Proceedings of the International Conference on Advances in ICT for Emerging Regions (ICTer), (2014 December)
Dantanarayana G. and Jayaweera P.,Framework to Capture, Represent and Realize Motivational Concepts of Good Governance on eBusiness Solutions, in Proceedings of the 1st International Research Conference and Social Sciences(IRCHS2012),(2012).
Dantanarayana G., Wickramage C and Jayaweera P., Value Object Classification for Healthcare Service Development, in Proceedings of the Peradeniya University International Research Sessions (iPURSE 2014), (2014 July)
Dantanarayana G., Wickramage C and Jayaweera P., Value Activity Categorization towards Healthcare Service identification, in Proceedings of the International Conference on Multidisciplinary Approaches (ICMA 2014), (2014 August)

UA Aruni Niroshika

Start Date: 01/01/2011

Completion Date: 21/10/2014

Email: aruni_niroshika(at)yahoo.com

Supervisor(s): Prof. RGN Meegama and Dr. Mahesha Kapurubandara

PhD Research Title: Active Contour based Medical Image Segmentation

The main contribution of this research study is to devise an effective active contour model to segment discontinuous boundaries in X-ray Images. This involves the improvement of the active contour’s capability in detecting sharp corners by incorporating prior knowledge about the significant corners of the object of interest, while preserving the general energy terms.

Active contours are a form of curves that deform according to an energy minimizing function and are widely used in computer vision and image processing applications. In this research study, a new technique of active contour model to extract boundaries of objects having sharp corners is devised. By incorporating a priori knowledge of significant corners of the object into the deforming contour, the proposed active contour is able to deform towards the boundaries of the object without surpassing the corners.

Moreover, all the significant corner points/discontinuous regions of the targeted object should be identified accurately during the initialization step of the proposed snake model. In order to perform this task with less user involvement, a competent corner detection operator should be integrated with the proposed snake model. Therefore a novel corner detection operator known as BLAID (Boundaries from Locally Adaptive Isotropic Detection), which is capable of detecting corner points that exist only on the boundary of an object, is integrated with the proposed snake model.

The ability of the new technique to accurately extract features of interest of anatomical structures in medical X-ray images having sharp corners is tested. The proposed model demonstrated significantly better results on detecting corners comparatively with the other existing models. We demonstrated the efficiency of our snake model by testing it on images with additive Gaussian noise. The experimental results demonstrate that the proposed snake model performs well for the images having an SNR of > 20.

Publications:

Journals

Niroshika UAA, Meegama RGN, Lokupitiya RS, Kannangara DKS. Active contours with prior corner detection to extract discontinuous boundaries of anatomical structures in X-ray images. IET Image Proc. pp. 202 – 210. Volume 9, issue 3, 2015.

Conferences

Niroshika UAA, Meegama RGN, Locally Adaptive Isotropic Detection of Corners in Object Boundaries, Proceedings of the International Conference on ICT for Emerging Regions IEEE; 2014 Dec; Colombo, Sri Lanka. (Accepted)
Niroshika UAA, Meegama RGN. A Deformable Model to Segment Discontinuous Boundaries of Bone Fractures in X-ray Images. Proceedings of the International Conference on Industrial and Information Systems (ICIIS 2013). IEEE; 2013 Dec; Kandy, Sri Lanka. p. 309-314.
Niroshika UAA, Meegama RGN, Fernando TGI. Active Contour Model to Extract Boundaries of Teeth in Dental X-Ray Images. Proceedings of the International Conference on Computer Science and Education (ICCSE 2013). IEEE; 2013 April; Colombo, Sri Lanka. p. 396-401.
Niroshika UAA, Meegama RGN. Active Contour Model for the Detection of Sharp Corners in Image Boundaries. Proceedings of the National Conference on Technology and Management (NCTM 2013). 2013 Feb; Malabe, Sri Lanka. p.9-13.
Niroshika UAA, Meegama RGN. Active Contours with Prior Corner Detection. Proceedings of the International Conference on Machine Vision, Image Processing, and Pattern Analysis (ICMVIPPA 2012). 2012 Feb; Kuala Lumpur, Malaysia. p. 682-686.

IDID Ariyasingha

Start Date: 01/10/2011

Email: iresha(at)ou.ac.lk

Supervisor(s): Dr. TGI Fernando

MPhil/PhD Research Title: Designing and developing Nature-Inspired Algorithms for Multi Objective Optimization Problems

In the nature, there are many intelligent social colonies. Researchers have extracted ideas from the nature to design nature-inspired algorithms to solving complex problems which cannot be solved by the usual techniques. Nature inspired algorithms are natural fit for solving problems in biological domains and also these algorithms search for new computational solutions to complex problems. Some of the nature-inspired algorithms are genetic algorithms, particle swarm optimization, ant colony algorithms and artificial bee colony. In the real world, there are many optimization problems which cannot be solved simply by a single objective. They are called multi objective problems. Nature-inspired algorithms can find optimal solutions in one simulation run due to their population-approach and can obtain solutions for large, complex and dynamic problems. Therefore, this research focuses to designing and developing nature inspired algorithms for solving multi-objective problems.

Publications:

Ariyasingha IDID, Fernando TGI (2015) Performance analysis of the multi-objective ant colony optimization algorithms for the traveling salesman problem, Swarm and Evolutionary Computation, doi: 10.1016/j.swevo.2015.02.003 [Abstract]
IDID Ariyasingha, TGI Fernando (2014) Analysis the Performances of the Multi-objective Ant Colony Optimization Algorithms, Proceedings of the 70th Annual Sessions of the Sri Lanka Association for the Advancement of Science, 503/E1. pp. 54.

MDR Perera

Start Date: 01/01/2012

Completion Date: 31/10/2014

Email: dilum(at)dscs.sjp.ac.lk

Supervisor(s): Prof. R.G.N. Meegama, Prof. M.M. Pathmalal and Dr. M. K. Jayananda

MPhil Research Title: Wireless sensor network for in-situ measurement of selected water quality parameters

In Sri Lanka, still we are unable to implement proper sensor network to measure water quality parameters such as pH, dissolved oxygen, electric conductivity and turbidity etc, because it requires high capital and operational cost.

In general, applications involving sensor networks operate in natural environments that exhibit dynamic behavior, it is often necessary to reconfigure the nodes to comply with the environment. Although network and sensor node failures may occur under a natural environment, a single node failure should not affect the entire network. It should be able to readjust communication within the network with remaining nodes and it should have the ability to add a new node to the network or remove an existing node from the network without affecting the whole system. Hence, the processing chip in the sensing modules must provide enhanced flexibility for the designer to implement a particular algorithm to accommodate this dynamic behavior.

Compact and low cost of each individual sensor node has a direct impact to the large scale deployments of the wireless sensor networks. Smaller nodes are easy to transport, possible to locate in limited space, able to use in different applications which has high mobility and maintenance too. Total budget of the sensor network application is calculated by considering the design, development, operational, logistics, implementation, and maintenance costs. Main potion of the budget covers the development and maintenance process. Therefore, the use of smaller, reliable, accurate and robust nodes will possibly reduce the overall budget of the project in the long run.

Those are the main fundamental issues we face in designing the process of the sensor network applications. In this project we mainly focused to address the basic fundamental issues to reduce initial and maintenance cost. For that we propose a single chip solution for interfacing transducers to reduce the cost per unit and use an IEEE 1451 family of standers in the implementation process to add intelligent capability to the node.

We design a sensor node including necessary functionality blocks such as data processing, transducer control, memory, communication protocol etc. in a single FPGA chip. As a result we able to reduce the number of interface chips.

FPGA is enhancing their capability by facilitating the dynamic re-configurability, which is able to modify the system design when it is under operation. Therefore, it provides high flexibility to design dynamic behavior systems than other chips. Furthermore, it helps to increases the performance significantly. That is the reason why we select a FPGA chip for this development.

IEEE 1451 family of standers provides implementation standards of TEDS that stores transducer identification, calibration, correction data, measurement range, and manufacturer-related information, etc. to facilitate common message and commands structure to connect transducers with systems or networks via a wired or wireless media. Through this implementation, we are able to add self- identification, self-diagnostics, and self-description capabilities to the proposed node.

Publications:

Journals

MDR Perera , RGN Meegama and MK Jayananda, “FPGA Based Single Chip Solution with 1-Wire Protocol for the Design of Smart Sensor Nodes”, Journal of Sensors, vol. 2014, Article ID 125874, 11 pages, 2014. doi:10.1155/2014/125874.

Conferences

MDR Perera, RGN Meegama and MK Jayananda, “Single Chip Solution with 1-Wire Communication Protocol to Interface Digital Transducers to Sensor Networks”, 36th Asia Pacific Advanced Networking (APAN) Research Group Workshop, Daejong, Korea, 2013.
MDR Perera , RGN Meegama and MK Jayananda, “A Single Chip Solution for Interfacing Transducers to Sensor Networks Using FPGAs”, 8th IEEE International Conference on Computer Science & Education , 2013, pp. 201.

RKAR KARIAPPER

Start Date: 01/10/2013

Email: kariyapper(at)gmail.com

Supervisor(s): Dr. Prasad M. Jayaweera and Prof. Hans Weigand

MPhil/PhD Research Title: Ontological driven legal service process management

The research is indented to develop opimized ontology framework legal process management

Publications:

Kariapper, R.K.A.R. and Jayaweera, Prasad M, Dynamic Essential Modeling of Organization (DEMO) Towards the Legal Domain (February 25, 2015). Available at SSRN: http://ssrn.com/abstract=2607921 or http://dx.doi.org/10.2139/ssrn.2607921
RKAR Kariapper, PM Jayaweera, A Sruvey on Legal Ontologies for the Development of Process Optimization Ontology, International Conference on Multidisciplinary Approaches (2014)
RKAR Ahmadh Rifai Kariapper, Prasad M Jayaweera, A Comparison of Selected Legal Ontologies for the Design of Legal Process Optimization Ontology, International Journal of Scientific Researches Vol-3 Issue-12, 2014
RKAR Kariapper, PM Jayaweera, Process Marting Approach for Legal Record Archiving, International Research Conference on Humanities and Social Sciences, 18 (2013)

MKA Ariyaratne

Start Date: 01/10/2013

Email: anuradhaariyaratne(at)gmail.com

Supervisor(s): Dr. TGI Fernando and Prof. Sunethra Weerakoon

MPhil/PhD Research Title: Designing and Developing Nature Inspired Algorithms to solve Nonlinear Equations (on going research project)

This research focuses on developing some nature inspired algorithms to solve a given set of nonlinear equations that are difficult to solve by classical approaches. Initially solving a single nonlinear equation having several real roots within a specified interval is considered. Then improvements can be made to find complex roots and the final goal is to solve a given system of nonlinear equations using Nature inspired algorithms. Latest nature inspired meta-heuristic algorithms like Firefly algorithm, Bat algorithm, cuckoo search algorithm will be used and some other old algorithms can be used for the comparison purpose.

Publications:

MKA Ariyaratne, TGI Fernando & S Weerakoon, Finding Roots of a Nonlinear Equation Using the Bat-Inspired Algorithm, Proceedings of the 70th Annual Sessions of the Sri Lanka Association for the Advancement of Science, 2014. 501/E1. pp. 52.

S. Vasanthapriyan

Start Date: 01/10/2013

Email: priyan@appsc.sab.ac.lk

Supervisor(s): Dr. Prasad M. Jayaweera and Prof. Hans Weigand

PhD Research Title: Unified framework for legal process repositories

The main focus of the research is to enable the district courts to incorporate the information technology within the legal system by:

Gather all the district court activities as process models
By using process modeling technique, develop the information bank which enhance the storing and retrieval of the information, which leads on-time verdicts to the public
These researches aim at developing frameworks or conceptual models

Publications:

Books:

Shanmuganathan Vasanthapriyan, R.P.T. Bhagya, and Prasad M. Jayaweera, “Collaboration Modelling Framework for Legal Workflow Management: Collaboration Modelling Framework for Courts Hearing Workflow Specification for Sri Lankan Context”, ISBN-10: 3659608475, ISBN-13: 978-3659608476, LAP LAMBERT Academic Publishing, 2014

Journal Articles:

T. Bhagya, S. Vasanthapriyan, and P. Jayaweera, “Meta-model for Collaboration Modeling in Legal Sector”, International Journal of Multidisciplinary Studies (IJMS), Volume 1, Issue 1, ISSN 2362-0797, Page 113-122, 2014
Thilini Bhagya, Shanmuganathan Vasanthapriyan, and Prasad Jayaweera, “Collaboration Modelling Framework for Courts Hearing Workflow Specification”, Journal of Developing Country Studies, Volume 4 – Number 13 (Special Issue), ISSN 2224-607X (Print), ISSN 2225-0565 (Online), Page 83-92, 2014
R.P.T. Bhagya, S. Vasanthapriyan, and P. Jayaweera, “Survey on Requirements and Approaches of Business Process Repositories”, International Journal of Scientific and Research Publications (IJSRP), Volume 4, Issue 1, ISSN 2250-3153, 2014

Conference Proceedings (Full Paper):

R.P.T. Bhagya, S. Vasanthapriyan, and P. Jayaweera, “Pattern Driven Collaboration Modeling in Legal Case Management”, Conference Proceedings – Part 4: Marketing Management and Technology, 5th International Conference on Business and Information (ICBI 2014), ISSN 2420-7497, Page 17-24, 2014

Conference Proceedings (Abstracts):

Thilini Bhagya, Vasanthapriyan Shanmuganathan, Prasad Jayaweera, “Collaboration Modelling Framework for Legal Workflow Management”, Abstracts of 4th Annual Research Session (ARS 2014), Sabaragamuwa University of Sri Lanka, ISBN 978-955-644-039-3, Page 30, 2014
T. Bhagya, S. Vasanthapriyan, and P. Jayaweera, “Meta-model for Collaboration Modeling in Legal Sector”, Conference Proceedings, International Conference on Multidisciplinary Approaches (ICMA 2014), ISSN 2386-1509, Page 85, 2014

M Punchimudiyanse Start Date: 01/01/2014 Email: mpunchimudiyanse(at)gmail.com, malinda(at)ou.ac.lk Supervisor(s): Prof. RGN Meegama MPhil/PhD Research Title: Real Time Translation of Sinhala Language to Sign Language

This research is aimed at developing a framework for real time processing of Sinhala language to Sinhala sign language. The proposed project focuses on providing real time conversion of voice or text to a group of hand gestures (via a virtual avatar) through a computer application.

Aurally handicapped Sinhala community is using Sinhala sign language (SSL) as their main communication medium with a normal person. To have a two way communication, one should understand hand movements of the deaf person who uses SSL (deaf person Normal person) and should be able to express his/her ideas to deaf person using hand movements in SSL (Normal person deaf person). The research is intended to solve the problem of communication from normal person’s side to a deaf person.

Proposed research recognizes the speech of a normal person whose speech is known to the system and output recognized Sinhala text in phase one. Sinhala text is translated to appropriate hand gestures of the SSL in the phase two. Research is carried on to handle conversational SSL (sign gestures exist matching Sinhala words) as well as Phonetic SSL (signs does not exist for Sinhala words Eg. Names) appropriately.

Third and final phase of the project is animating sign gestures using a PC based virtual avatar. Speech recognition, translation and gesture sequencing has to be performed asynchronously while animation of signs in progress in a suitable speed which is comprehendible to the aurally handicapped person. Publications:

M Punchimudiyanse, RGN Meegama, “3D Animation framework for sign language”, International Conference on Engineering and Technology, March 17-18, 2015, Colombo, Sri Lanka.

Priyanga Dilini Talagala Start Date: 01/07/2014 Email: priyangad(at)uom.lk Supervisor(s): Dr. TGI Fernando and Prof. RGN Meegama

MPhil/PhD Research Title: Human Age Classification Using Facial Images: Deep Architecture Approach (on going research project)

Human automatics age prediction has become an important research topic nowadays due to its wide range of possible real-world applications such as effective human-computer interaction systems, security control and surveillance monitoring etc. However, compared to other facial variations, the automatic age prediction has rarely been explored due to its more complicated nature and the inability of performing the task quite accurately with the help of existing shallow architectures.

During the recent past Deep architectures have given very promising results for some complex AI problems such as gender classification, human action recognition, hand-written digit recognition etc.. This study is therefore, in an attempt to find whether deep architectures can assist in improving computer’s age prediction ability as well than it was capable of doing the task with the help of shallow architectures. The outcomes of this research can be expected to have a significant impact to outside the scope of the thesis and in many areas other than age classification.

Publications: on progress

Ishan Sabar

Start Date: 01/10/2014.

Email: ishan.res(at)gmail.com

Supervisor(s): Dr. Prasad M. Jayaweera and Dr. Ananda Edirisuriya

PhD/MPhil Research Title: Enhanced HL7-SAIF Semantic Interoperability for a Unified Healthcare Service Framework

This research focuses on using Ontological Modelling approaches for the improvement and enhancement of Semantic Interoprerability for HL7-SAIF. HL7-SAIF is the predominantly used specifications generation standard in the IT-driven healthcare sector worldwide, and is designed to foster and promote Semantic Interoperability amongst participating Healthcare Providers or enterprises.The research would employ a ground-up, unified approach in the achievement of the said goal.

This research would go further and also look at the saliency and viability of employing the HL7-SAIF standard to improve the healthcare sector in Sri Lanka, looking into all aspects related to its introduction, pertinent socio-economic issues, and benefits to be derived from its use in the local context.

Publications: Under Progress.

TMKK Jinasena

Start Date: 01/11/2014

Email: kasunkosala(at)yahoo.com

Supervisor(s): Prof. RGN Meegama, Dr. RB Marasinghe

MPhil/PhD Research Title: Secure Mobile Collaboration Platform for Transmission and Archival of Medical Images

This research focuses on to design and develops a secure mechanism to transmit medical image such as MRI images of the brain, efficiently and effectively in a collaborative mobile environment. However the proposed secure mechanism would allow medical practitioners to multicast medical images with multiple accessibility levels over the content.

Publications : on progress

WPJ Pemarathne

Start Date: 01/01/2016

Email: punsisi(at)kdu.ac.lk

Supervisor(s): Dr. TGI Fernando

MPhil/PhD Research Title: Electrical cable optimization system for single story building using ant colony optimization

Solutions from the nature are one of the greatest ways to address the issues in the current world. Real world optimization problems are very hard to solve. One of the more efficient ways to solve these problems is to use nature inspired algorithms. There are several nature inspired algorithms which are very popular due to their high efficiency, such as genetic algorithms, particle swarm optimization, cuckoo search, firefly algorithm and ant colony algorithm. Out from this ant colony algorithm which represents the behavior of ants is a very special algorithm. Scientists have studied the behavior pattern of ants to solve many real world problems. This research is focus on developing the ant colony algorithm to design an optimized electrical cabling layout for a building.

Publications: on progress