Centre for Distributed Systems and Software Engineering...

www.infotech.monash.edu/dsse

Centre for Distributed Systems andCentre for Distributed Systems andSoftware Engineering (DSSE)Software Engineering (DSSE)

From theoretical foundationsto practical tools, solutions, services,applications, platformsand systems

www.infotech.monash.edu/dsse2

Centre for DSSE

• Director: A/Prof Arkady Zaslavsky• Deputy Director: Dr Shonali Krishnaswamy

• DSSE carries out research in foundations, methods, tools and environments for conceptualising, modelling, analysing, developing, maintaining and benchmarking large parallel, distributed and pervasive systems.

• Established in 1998 as Monash University Research Centre• Ranked Faculty Research Strength #1

– 1998, 1999, 2000, 2002– Weight in Caulfield & Clayton Schools of IT – Incl. strengths in Gippsland School of IT

• Strong international collaboration and exchange program• Members and other details see

– http://www.dsse.monash.edu.au/


DSSE mission

• Mission: DSSE aims to advance the foundations, methods and practice of distributed systems and software engineering. Its mission is to achieve internationally recognised research excellencethrough national and international competitive grants, high impact and quality publications, extensive international collaboration, product-oriented industry collaboration, focused research student and staff exchange and efficient and effective PhD student supervision.


DSSE structure

• MESSAGE Lab (E-Science and Grid Engineering Laboratory)– Director: Prof David Abramson

• Mobility Lab (Mobile, Ubiquitous, Pervasive Computing and Databases)– Directors: A/Prof Arkady Zaslavsky & Prof

Bala Srinivasan


MESSAGE Lab (E-Science and Grid Engineering Laboratory)

• E-Science, enabled by the emerging Grid computing paradigm, tightly couples scientists, their instruments, massive storage devices and powerful computational devices. The computational Grid has promised a great deal in support of innovative applications, particularly in science and engineering, but developing applications can be demanding.

• The Lab is developing a range of upper middleware services and tools to solve the software engineering challenges in building real applications.

• Current research includes: (a) tools that make it simple to build Grid applications, leveraging existing legacy codes, (b) systems for testing applications on the Grid, (c) systems to deploy application to the distributed heterogeneous environment and (d) application execution mechanisms, including run time environments for distributed workflows.


Mobility Laboratory

• The Mobility Laboratory carries out research in pervasive, mobile, ubiquitous and wireless computing and networking.

• A diverse range of projects investigates the technical and social challenges of creating distributed computational environments and systems that will recede into the background of our everyday lives and empower us to be more effective in all of the activities that we undertake.

• Pervasive computing goes beyond the realm of personal computers. Almost any device, from clothing to tools, to appliances, to cars, to homes, to the human body, to your coffee mug, can be embedded with chips to connect it to an infinite network of other devices. The goal of pervasive computing, which combines current network technologies with wireless connectivity, Internet capability, artificial intelligence and other enabling paradigms, is to create an environment where the connectivity of devices is embedded in ways that are unobtrusive and always available.


Initiatives, spin-offs & research networks

• CRC DSTC – Distributed Systems Technology Centre, 1998-2005– M3, TrustMe, Nimrod

• MC2 Lab (Monash Cluster Computing Lab), 2001-2006– Joint: DSSE, CSSE, Maths, Earth Science– SMURF $150K + $120K

• CoolCampus - > $1.3M in-cash & in-kind, 2002-2006– Faculty of IT, Library, ITS, CeLTS, Student and Staff Services, other

faculties– 12 academics (CSSE, SIMS, SNC), 20+ student projects, 8 PhD use

infrastructure• M-commerce – CSSE & SIMS, 2000-2005• Links with faculties of Engineering & Science• EII (Enterprise Information Infrastructure) – ARC Research Network• ARC Research Network - National Molecular and Materials Structure

Network (MMSN).


International Context

• Extensive international collaboration network– DSSE researchers also have had close collaborations with several

Universities internationally including, UC San Diego, US; IRISA/INRIA Rennes, France; Luleå, KTH and Mäladarlen in Sweden; Fernuni Hagen, Germany; University Karlsruhe, Germany; University Oldenburg, Germany; Kings and Imperial College in London plus Cranfield University and the Welsh Grid centre in Cardiff, Technical University of Berlin, Germany, University of Pittsburgh, USA; NTNU, Norway, Oxford University, UK, University of Amsterdam, Netherlands, University of Linz, Austria, National University ofSingapore.

– Key members have attracted a significant amount of internationalgrants from multi-national industries.

• Intensive program of International visitors– With modest funding, the Centre has supported a vibrant international

visitors program hosting international research exchanges and sabbatical visits that have contributed significantly to the research context for PhD students and staff alike.


Composition of DSSE (20.08.2007)

• Number of RG members: 12• Number of Research Only Staff: 6• Number of Early Career Researchers: 3• Number of Associates: 7• Total Outputs* (2000-2006): 439

* Measurable outcomes, eg, publications, software, etc


Grants & income (2000-2005)

• ARC: 12 - $772,401• Government: 2 - $66,128• Industry: 14 - $577,634• International: 8 - $1,427,036• CRC: > $4,000,000


HDR (2000 – 2007) & Graduate Studies

• HDR completions: > 50• HDR students supervised (20.08.2007) : 53

DSSE directly and indirectly coordinates and delivers four forthcoming Master of ICT specializations (commencing 2009):

• Distributed and mobile computing systems• Data management• Internet and web applications development• Software engineering


DSSE research strengths

• GRID tools• e-Science• parallel computation• software engineering • high-performance

computational infrastructure

• middleware,• tools & platforms• parallel and relative

debugging• cluster computing• distributed workflows

• wireless solutions• mobile applications• multi-agent systems,• context-aware intelligent

environments• pervasive computing

systems• distributed data

management• sensor networks,• service-oriented

architectures• mobile networks• ubiquitous computing• security solutions

www.infotech.monash.edu/dsse

Centre for Distributed Systems andCentre for Distributed Systems andSoftware Engineering (DSSE)Software Engineering (DSSE)

Researchers and projects


Professor David Abramson

• New Methods in Scientific discovery – e-Science & e-Research

• Computational Platforms– The Grid and the Web

• Supporting a Software Lifecycle• Grid Services & Middleware• Software Lifecycle Tools

– Applications development– Deployment– Test and debugging– Execution

• Monash Tools– The Nimrod Family

> Applications– Deployment tools– Active Data– More …

http://www.csse.monash.edu.au/~davida


MiddlewareGlobus GT4 CondorAPST

Environmental Sciences

Life & Pharmaceutical

Sciences

ApplicationsGeo Sciences

Courtesy IBM,Lower Middleware

Upper Middleware & Tools

Bonds

PlatformInfrastructure Unix Windows JVM TCP/IP MPI .Net Runtime .Web Services VPN SSH


PlatformInfrastructure Unix Windows JVM TCP/IP MPI .Net Runtime

Environmental Sciences

Life & Pharmaceutical

Sciences

ApplicationsGeo Sciences

VPN SSH

LowerMiddleware Globus GT4

UpperMiddleware/Tools

Nimrod/G /O /E Kepler

Nimrod/K

Development

Debug

Test-Debug

Worqbench

IE EclipseEclipse

Guard

Eclipse VS

Deploy Motor

DistANT

Deployment

REMUS Active Data

NimrodPortal& WS

GriddLeSNetFiles Active

Sheets

Excel

Execution

SRB


• Supports workflows for robust design and search– Vary parameters– Execute programs– Copy data in and out

• Sequential and parallel dependencies• Computational economy drives scheduling• Computation scheduled near data when

appropriate• Use distributed high performance platforms• Upper middleware broker for resources

discovery• Wide Community adoption

Nimrod

1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006

Nimrod/GEnFuzion (www.axceleon.com)

Nimrod/ONimrod/OI

Nimrod/K

Active Sheets (Excel)

Nimrod Roadmap

Nimrod/WS


Nimrod/G

Grid Middleware

Nimrod/O Nimrod/E

Nimrod Portal

Actuators

The Nimrod family


Professor Bala Srinivasan

• Mobile computing• Wireless and mobile databases• Security in wireless and sensor

networks• Payment Systems and Security• Multimedia Retrieval Systems• QoS and QBiz in Networks and

Web Services

http://www.csse.monash.edu.au/~srini/


Data Broadcasting for Wireless Databases

Server

Data Broadcast Management

(Proposed Schemes)

Broadcast Channel

Listening to the channel

Retrieving the desired data

Data Repository

Retrieve database items

Required database items

MobileClients

Transmitter

Determining Optimum Number of Broadcast

Channels

Formulating Data

Broadcast Ordering and Scheduling Schemes

Index Broadcasting Schemes

• Traditional Queries

• Location-Dependent Queries

Minimising Query Access

Time

Minimising Tuning Time and

Power Consumption


Mobile Query Processing IncorporatingMultiple Non-collaborative Servers

Process data from twoDifferent servers- how to process?


Sensitive Information Retrieval Systems


Denial of service in Sensor Networks

Non-participating nodes

Gateway

Victim

Attackers

Non-participating nodes

Victim

Attackers

Laptop-class Adversary

DDoS attack launched against a victim

in a Wireless Sensor NetworkDDoS attack against a victim in an HPN environment


A/Professor Arkady Zaslavsky

• Mobile, wireless, pervasive, ubiquitous computing and networking

• Context-awareness• Mobile databases &

transaction management• Applications of pervasive

computing technology• Mobile software agents• Distributed computing

systems

http://www.csse.monash.edu.au/~azaslavs


LTU-Monash collaboration – PhD-polis

LULEÅTEKNISKA

UNIVERSITET

Adaptive & ContextAdaptive & Context--aware aware Multimedia in adMultimedia in ad--hoc communitieshoc communities

KKåårere SynnesSynnes

Christer Åhlund

Arkady Zaslavsky

Robert Brannström– multiple gateway

connectivity

Ruwini Kodikara –Networking issues, cross-layer context

exchange

Arkady Zaslavsky


Some Current projects

• Mobile transactions and web-services – SunamPradhan (ARC Linkage with RMIT)

• ARC Linkage with Intel – pervasive computing & crisis medicine

• ARC RN EII – taskforce on context-awareness• Robocollectors in sensor networks – Prem

Jayaraman – with LTU, Sweden• Data stream mining visualisation – Brett Gillick• Querying data streams – Pari Delir• 3 new PhD students have joined in 2007


Context-Situation pyramid

Situations

Context

Sensory-originateddata


28

Context Service

The architecture of a context service


Agent Injection Service

Load Monitor

Reasoning Engine

Cleaner

KB

Data AssimilatorData AssimilatorData Assimilator

Data InserterData InserterData Inserter

<raise process>

<raise process>

Monitor

<inject>

Stability Analyser

1. Knowledge Synthesiser

2. Conflict Analyser

3. Context Adapter

4. Verifier

MQ

MQ

SensorSensors

AgentFiltering Agent

Reasoning AgentAgent

Agencies

Event Router

Reasoning Layer

Algorithms Library

ContextExplorers – reasoning about context

Agent Agent communication communication platform using platform using

publishpublish--subscribesubscribe

ContextExplorersContextExplorers

A.PadovitzA.Padovitz, , A.ZaslavskyA.Zaslavsky & & S.W.LokeS.W.Loke, , B.BurgB.Burg, , C.BartoliniC.Bartolini


ITAG – Itinerary Agents

• ITAG – Itinerary Agents – mobile agents

– A high-level language for reasoning with and programming the mobility behaviour of agents


E-Hermes: Context-rich mobile agent technology to support information needs of financial institutions

e-Hermes

Mission Control Centre

Actor

Ontology

Mission Agent Factory(AF)

AgentComponent

Library

Mission Control Agent(MCA)

M

M

M

Log

Ontolo

gy

Mission Manager

ProfileRepository

Filter/Presentation toUser

Request

StaticInformation

Resulting informationResult

User Profile

MissionRepository

RequestInformation

User InteractionProcessor

Profile Builder

Glenn Glenn JayaputeraJayaputera ––AgentsAgents--onon--Demand Demand

Oshadi Alahakoon Oshadi Alahakoon –– PersonalisationPersonalisation & &

profilingprofiling

ARC Linkage, 2003ARC Linkage, 2003--20052005


SoundHunters: Mobile Agents Using Sound

• Track a moving sound source by listening to the sounds it emits and migrate through the network in order to always stay close to it, at least a step ahead

• Grasshopper, IBM ViaVoice, Speech Recognition, Sound Files

• Media attention !!!

X

X

X

(a) (b)

(c) (d)

Node with an agentNode without an agentA removed agentDirection of a copySound source

Key

X


Context-aware mobile real-time decision support for contingency management

• Team: F.Burstein (FIT), A.Zaslavsky (FIT), L.Churilov(BusEco), 1 medical researcher (TBA)

• The Research Aim is to propose an approach to sensitivity analysis to predict consequences of changing Quality of data (QoD) attributes, and analysing the impact of these changes to expected decision outcomes;

• develop a prototype mobile contingency management context-aware DSS, which is capable of providing QoDinformation in a user-specific way.

• establish QoD metrics that will take into account static and dynamic properties of the mobile environment;


CoolCampus Ideology

Pervasive/Ubiquitous/Mobile Research•Location awareness•Sensor networks•Adaptability•Context management•Resource constraints•Devices & services•Software•….

University users•Administration•Library•Student services•Academic board•IT Services•Community sector•Monash International•…..


CoolCampus’2003: Projects…

• SoundHunters: Mobile Agents Using Sound• Location-aware book finder• Mutually location-aware mobile robots• Enabling A Smart Seminar Room Using Mobile

Agents• Location Awareness & Sensor Networks• CIM: Cool-campus IPAQ Monitor• Mobile phone based personal assistant


CoolCampus’2004: Projects…

• Commercially Available Sensor Devices for Location Aware Infrastructure

• Wireless network topology visualisation

• A System for Proactive Discovery and Update of Location-Aware Ambient Services

• Hanging Services: An Investigation of Context-Sensitivity and Mobile Code for Localised Services

• ContextExplorers


A/Professor John Hurst

• Document technologies• Educational technologies• XML and Literate

Programming• Expert Student Course

Advice

http://www.csse.monash.edu.au/~ajh


XML and Literate Programminghttp://www.csse.monash.edu.au/~ajh/research/literate/index.xml

• Literate Programming is a technique developed by Donald Knuth, and described in his seminal paper in Computer Journal, 1984. Itis both a software process and a suite of tools for supporting that process. Programs written as literate programs go through two (conceptually) independent processes known as tangling and weaving. Tangling creates the actual source code for input to a compiler; weaving creates a documentation file suitable for processing by a document system such as TeX.

• More modern versions of literate program tools have attempted todecouple the N x M combinatorial explosion created by the choiceof N programming languages and M documentation systems, by restricting the markup and presentational mechanisms. The tool currently in use in the work described here is such a tool: the markup of source documents is done in XML, and the documentation device is handled by XML translators. This reduces the problem from an N x M to an N + M one, where N = 1.


Expert Student Course Advice

• As part of an on-going process to better utilize the advantages of information technology in a university learning environment, a recent student project looked at ways of extracting course and unit information from publically available handbooks, and turning this into an "expert system" that could provide course advice tostudents. One outcome from this project was the automatic rendition of a Prolog program that was tailored to a given student at some arbitrary point in his or her course. Knowing what units had been completed, what units the course structure required, and what units might be undertaken in the next semester, the program would not only supply the student with a list of possible units that could be taken over the next semester, but also would show what units had to be taken in order to complete the course with (say) a specialization in a particular field.

• The project continues to develop a workable prototype. This prototype could be employed via say a web interface, or faculty kiosk."


Dr Shonali Krishnaswamy

• Distributed and Ubiquitous Data Stream Mining

• Web Services and Service Oriented Computing

• Intelligent, Mobile Software Agents

http://www.csse.monash.edu.au/~shonali


Current/Recent External Research Grants

• Taskforce on Context-Aware Computing (2007)– Role: Taskforce Leader– Funded By: ARC Research Network on Enterprise

Information Integration (EII)

• ARC Linkage - Ubiquitous Data Mining and Situation-Awareness for Improving Road Safety (2005-2007)

– Role: Chief Investigator (with A/Prof Rakotonirainy and Prof M Sheehan (CARRS-Q, QUT) and Dr S W Loke (La Trobe Univ.)

– Industry Partner: Insurance Australia Group

• ARC Discovery - Efficient Prediction of Application Metrics for E-Services (2003-2005)

– Role: Australian Post Doctoral (APD) Fellow (with A/Prof A Zaslavsky as CI)


Research Projects – Ubiquitous Data Stream Mining

• Architectures– RA-UDM (Resource-Aware UDM)– Context-Aware Middleware for Adaptive Data Stream Mining

• Algorithms– Lightweight Clustering (LWC) and a Fuzzy extension for interpretation/labelling– Lightweight Classification (LWClass)– Lightweight Frequent Items (LWF)– RA-VFKM – Resource-Aware Very Fast K-Means– RA – Cluster - focuses on adaptation at input, processing and output points – Resource-Aware and Adaptive Visualisation Techniques for Handheld Devices

• Applications– Intelligent Transportation Systems - Intersection Collision Detection and Warning,

Accident Prevention at Curves, Real-Time Detection of Drunk Driving Behaviour (with CARRS-Q @QUT and Insurance Australia Group)

– Sensor Networks – Context-Awareness and In-Network Mining for Energy Conservation in Sensor Networks (with Department of Primary Industries, Victoria)


Resource-Aware UDM (RA-UDM)

Adaptation of learning algorithms based on computational context and stream characteristics - Algorithm Output Granularity (AOG)


Context-Aware Middleware for Data Streams

Adjustable adaptation strategies based on context and situation changes –of user, device, connectivity and application


Research Projects – Web Services

• Web Services Reputation– Verity: a new QoS metric to be used in conjunction with

reputation– Rationale Analysis: models for inferring rationale of web

services ratings, determine collusion, deception as well as make personalised recommendations

• Web Services QoS– Predicting QoS Levels to Provide Differentiated Services– Pareto-Optimal Scheduler to Improve Quality of Business

(QoBiz) and Quality of Experience (QoE)• Pervasive Web Services

– Pervasive Service Discovery with Context and Semantics– Specifying, modeling and verifying context in service

contracts or service level agreements


Research Projects - Intelligent Mobile Software Agents

• AMM – Agent Mobility Modelling

• CALMA – Context-Aware Mobile BDI Agents

• RAMMA – Resource-Aware Middleware for Mobile Agents

• BlueM – Mobile Agents over Bluetooth


Dr. Chris Ling & Dr Maria Indrawan

• Main Research Interests: – Software Engineering, Specification languages, – Petri nets, Workflow analysis– Context-aware systems, Pervasive computing

• Two current projects:1. Device Ecology Workflow2. A Model of Mutual Awareness for Context-Aware Systems.

http://www.csse.monash.edu.au/~sling/ http://www.csse.monash.edu.au/~maria


Device Ecology Project

• Supported by ARC (Australian Research Council) Discovery (2004-2006)

• To develop a framework for intelligent, mobile and context-aware devices to integrate, communicate and interact with each other.

• Device ecology:– A collection of devices (in the environment and on

users) interacting synergistically with one another, with users and with Internet resources, undergirdedby appropriate software and communicating across the environment.


Device Ecology Workflow

• Device ecology encompasses device aggregation, service-oriented computing and workflow management studies.

• Decoflow (Device Ecology Workflow) - As a metaphor for thinking how collections of these devices can work together to accomplish a purpose.

• Web services as the communication channel– WSDL - defines services – BPEL4WS - coordinates services

• Decoflow engine which executed BPEL specification.


Architecture

User Interface Level

Translator

Device Conversations Execution and Management

Device Ecology Workflow Engine

Device Ecology and Web Resources

User Inputs

High-level workflow (ECO Expressions)

BPEL4WS Specifications

Conversations

Service Calls


Device Ecology Projects

• Students: 5 Master of IT (Minor Thesis) and 2 BComp(Honours).

• Project titles:– Analysis of comfort levels in a device ecology workflow.– Dynamic changes in device ecology workflow– Translating high-level workflow language to BPEL4WS– A synchronization strategy for device ecology workflows– Determining non-determinisms in workflow executions

based on contexts– Towards RFID-based user profiling– Adaptive Workflow for Devices based on context


A Model for Mutual Awareness for Context-Aware Systems

• Interaction among devices based on Spatial Model of Interaction (Benford and Fahlen, 1993)

• Abstract concepts: Aura, Focus and Nimbus:– An Object A with a higher value or level of Focus over the

value or level of Nimbus of another Object B will be aware of Object B.

– An Object A with a higher value or level of Nimbus over the value or level of Focus of another Object B, will not allow Object B to be aware of Object A.

• To realise the abstract concepts to the real environment.


Realising Aura for Initiating Interactions in Real Environments

• There are three considerations when mapping the abstract concepts into real environments:

– Aura Types> Needed to determine what type of interaction is to occur.

– Levels of Granularity in Modelling Auras> Affects the amount of control given to users for interactions

and services.– Collision Models

> Affects how collisions are detected in real environments.

• Involves 2 BComp (Honours) students and 1 MIT (minor thesis) student.


Visualization of Concepts


Aura Application


Communication Model


Dr Sita Ramakrishnan

• Research interests– Software engineering– Validation of distributed software components – Compliance and conformance testing of

software components – Validation of software testing tools – Concurrent and distributed systems and

software engineering – Component-based architectures and

development – O-O method, process, quality, metrics and

testing – Software engineering and web engineering – Web technologies in education and research

http://www.csse.monash.edu.au/~sitar/


Virtual Research Lab: Monash University Software Engineering Services Architecture and Testing (MUSES-AT) Lab

• Component-based software engineering and software architectures have become a hallmark for distributed systems software development. Current research includes:

– service-oriented architectures for scalable real-time distributed systems and sensor grids

– design testable service-oriented software systems for sustainable manufacturing

– reliability, trustworthiness and context-awareness– architecture-driven modelling, model-driven verification, testing and

monitoring– aspect-orientation in component-based software– Identity Management

• Dr Sita Ramakrishnan, Clayton School of IT• Doctoral Students: Ms Nor Laily Hashim, Mr Dhaminda

Abeywickrama, Ms Susan Entwistle, Mrs Sreechandana Madiraju, Mr Daniel Rowley, Mr Quazi Mamun


Dr Iqbal Gondal

• Research interests– Networking and

Communication – Information fusion – Ad hoc network networks – Gene Networks – On-line education – Technology transfer

http://www.gscit.monash.edu.au/gscitweb/index.php?type=i&id=44


List of current research projects/activities, including any ARC grants

• Projects:– Sensor ad hoc networks for Cardiogenome monitoring of

cardiac patients– Self configuration for ad hoc sensor networks– Adaptive drug discovery monitoring using ad hoc networks– Vertical soft hand over for next generation wireless networks– Smart antenna for ad hoc networks– Genomic information fusion for post genomic knowledge

discovery– Sensor and actor networks for nursing homes

• ARC– Adaptive drug discovery monitoring using sensor ad hoc and

gene networks (800k) - pending


Tailored Congenital Heart Patient Monitoring

The project would enable medical services providers to provide tailored, critical and cost effective care to the cardiac patients by utilizing their DNA profile to determine mutation levels in their selected genes. The intelligence of pervasive network and cardiac genome servers would ensure patient safety and continuous monitoring and at the same time would reduce false alarm for cardiac patients.


Dr David Taniar

• Research interests:– XML databases– Object relational databases– Mobile databases– High performance parallel

databases– Grid databases– Data warehousing and

mining

http://users.monash.edu.au/~dtaniar/


XML and Object-Relational Databases

• XML is a semi-structured data growingly used for data exchange on the web.

• Research areas:– XML query processing– XML storage and updates– XML warehousing

• Collaborators: W.Rahayu (La Trobe) and S.Bressan(NUS)

• PhD graduate: E.Pardede (La Trobe, 2007)• Book: Rahayu, Taniar, and Pardede, Object-Oriented

Oracle, 2006.


Mobile Databases

• Mobile data access gives mobile devices capabilities to access data through wireless communications

• Research areas:– Mobile query processing– Mobile data broadcast

• PhD graduates: A.B.Waluyo (Monash, 2006)• Book: Taniar, Waluyo, and Srinivasan, Mobile Query

Processing, John Wiley & Sons, 2008.• Editorship: Mobile Information Systems (IOS Press,

The Netherlands)


High Performance Parallel/Grid Databases

• Parallel/grid database processing makes use of parallel and gridcomputing in database access to achieve high performance.

• Research areas:– Parallel query processing– Grid databases– Semantic grid

• Collaborators: C.H.C.Leung (VU), W.Rahayu (La Trobe), and B.O.Apduhan(Japan)

• PhD graduates: A.Flahive (La Trobe, 2007), S.Goel (RMIT, 2005), and L.Tan (Monash, 2005)

• Book: Taniar, Leung, Rahayu and Goel, High Performance Database Processing and Grid Databases, John Wiley & Sons, 2008.


• Early career researcher• Primary Research Interests

– Multimedia databases and multimedia retrieval systems

– Multimedia data mining– Quality of service in

networked multimedia delivery

Dr Campbell Wilson

http://beast.csse.monash.edu.au/~cwilson/wp/


Current Research Projects (PhD)

• Samar ZutshiRelevance Feedback in Multimedia Systems

• Klangjai SithitavornQuality of Service Routing in MANETS

• Mylini MunusamyAdaptive Web Personalisation

• Gita Das• Low Level Feature Analysis in Content-based Image

Retrieval • Resmi Nair

Query-driven Conceptual Architectures for Data Warehouses (Thesis submitted)

• Harry NgoMobile Computing Authentication Models

• Noor Azilah DramanMetadata Representations for Content-based Audio Retrieval


Relevance Feedback in Multimedia Systems

• With PhD Student Samar Zutshi • Recognise the utility of decoupling the relevance

feedback engine of content based retrieval systems from the retrieval engine. Model relevance feedback as a separate classificatory problem.

• This separation allows the feedback to be sent to multiple retrieval engines and also allow feedback on the retrieval engine itself.

• Suggest optimum retrieval strategies within a given model based on rough set analysis of user feedback.

• Within the framework, establish system-centric criteria for evaluation of retrieval results which can be used to suggest relevant multimedia databases to particular classes of user.


Adaptive Web Personalisation

• With PhD Student Mylini Munusamy• Restructuring of web graphs as an optimisation problem.• Utilise syntactic (structural) properties of the graph in

addition to harvested behavioural attributes of users.• Behavioural attributes gathered on server and client side.• Content based attributes (e.g. tf/idf values)• How can learning algorithms be employed to suggest

web restructuring based on harvested attributes?• Extend to include attributes describing end user

browsing devices.• Goal is device/connection-adaptive and personalized web

site construction algorithms and designs.


QUESTIONS ?

The EndThe EndThanks for browsingThanks for browsing

Centre for Distributed Systems and Software Engineering...

Documents

Transcript of Centre for Distributed Systems and Software Engineering...