Research Projects
Available research projects for 2008 are listed below under the Faculty's main research themes (including research collaborative partners):
eResearch
Business Services
Information Science
Systems Science
Information Security Institute
CRC Smart Services
For information on applying for admission into a research course click here
Research Area: e-Research
Smart Tools Project
Lead Investigator Associate Professor Jim Hogan
The smart tools project has a large interdisciplinary team building next generation scientific workflows and supporting their application to key biological questions - notably understanding regulation in both bacteria and higher organisms. Our work involves close collaboration with local and international biologists, with interests in infectious diseases and cancer.
PhD opportunities are available for highly motivated students with a strong background in one or more of Software Technology, Workflows, Machine Learning and Bioinformatics. All of our projects are geared towards answering some of the following questions:
• How can workflows (virtual scientific experiments) be naturally specified and implemented?
• How can large comparative studies be undertaken without undue pain to the users?
• How can data from high throughput sequencing and scanning machines be filtered to enable analyses on what would otherwise be infeasibly large datastreams?
• How can workflows, data and publication results be described in order to support automation and interactions within the scientific community?
• How can scientific discovery be semi-automated?
Medical Image Analysis for Early Cancer Detection
Lead Investigators Professor Binh Pham, Professor Paul Roe, Associate Professor Jim Hogan, Dr. Jinglan Zhang, Dr. Jinhai Cai
Prostate cancer is the most common form of cancer in men worldwide with approximately 25% of male cancer diagnoses being prostate cancer. This project will investigate image processing and pattern recognition techniques for efficient and effective medical image classification and recognition. We aim to develop new algorithms that will allow more accurate and objective diagnoses of prostate cancers using a clinical image database by recognising textual features.
Acoustic Information Analysis and Retrieval
Lead Investigators Dr Jinhai Cai, Dr Jinglan Zhang , Prof Binh Pham
This project will investigate algorithms for bird species recognition based on bird calls and includes machine learning and acoustic signal processing algorithms. Data is collected at the Samford Ecological Research Facility established by the eResearch Centre in 2007.
Image Analysis and Retrieval for Environment Monitoring
Lead Investigators Professor Binh Pham, Dr Jinglan Zhang , Dr Jinhai Cai
This project will investigate algorithms for bird species recognition based on bird images including machine learning and image processing algorithms.
Build 3D Models from 2D Images
Lead Investigator: Dr. Jinhai Cai
Building 3D models from 2D images has many applications in visualisation, medical image processing and vegetation management. The project will employ techniques from stereo matching or optical flow to build 3D models from 2D images taken by cameras in UAVs, cars or airplanes. You may work with people working on CRC-SI project.
Acoustic Clues for Video Retrieval
Lead Investigator: Dr. Jinhai Cai
To build a real-time system for video retrieval is very challenging. This is because image processing is a time consuming task. However, acoustic signal processing can be 10~30 times faster than real time. It is useful to retrieve interesting segments from videos using acoustic signals. The goal of this research is to find video segments with voices or music. It can be applied in sports, security and surveillance. You may work with people from Microsoft/QUT eResearch centre.
Robust Fingerprint Recognition
Lead Investigator: Dr. Jinhai Cai
Fingerprint recognition has many applications in security, and surveillance and forensic sectors. As the quality of fingerprints are not necessary high, we have to develop new techniques that are robust to noise in order to reliably and accurately recognise these fingerprints. The goal of this project is to develop techniques that are able to enhance the quality of fingerprint images sampled from scenes, to determine reliable regions of a given fingerprint image and extract critical features, and finally to identify the person.
Recognising and Tracking Objects from Infrared Videos
Lead Investigator: Dr. Jinhai Cai
Infrared cameras are widely used in security and surveillance systems. It is desirable to build a system that is able to recognise and track objects from infrared videos and further issue warning of danger, instead of just recording images in current systems.
Reasoning about Parallelism
Lead Investigator Wayne Kelly
When we buy a new computer we expect it to be considerably faster than our old computer. CPU performance has been doubling approximately every 18 months - a trend that has been consistent for the last 30 years. Unfortunately, fundamental limits of physics means that this will soon come to an end. In the future, if we are to continue to see increased performance, it will come about primarily by exploiting parallelism. We are already seeing this in the widespread release of multicore chips. However, for an individual application to exploit this parallelism we need to be able to reason about the inherent parallelism present. We do not necessarily want programmers to write explicitly parallel algorithms, but the languages that we typically use to express algorithms today makes it extremely difficult for either programmers or compilers to reason about parallelism. This project looks to develop new programming paradigms and analysis techniques that will aid in reasoning about parallelism. In a related project, we are investigating how best to exploit emerging parallel hardware such as Multicore chips and how to utilize graphics processing units (GPUs) as general purpose processing units.
Information Visualization
Lead Investigators Professor Binh Pham, Dr Ross Brown, Dr Jinglan Zhang
Visualization is the merging of abstract and symbolic data with the display of concrete geometric objects through computer graphics (modelling, transforming, shading, lighting, and animating) in order to communicate the information in a dataset to enhance understanding. Visualization seeks to determine and present underlying correlated structures and relationships in both scientific (computational and medical sciences) and other application datasets (survey, census etc).
The research in this group aims to use computer-supported tools to make information visible so that it can be better identified and understood in order to support effective technical and business decision-making. The research will be conducted at different levels including theories, methodologies, techniques, applications, evaluations and case studies in relation to information visualization.
Browsing the Physical World in Real-Time using Sensors and SensorMap
Lead Investigators Professor Paul Roe, Professor Binh Pham, Dr Jinglan Zhang
Geo-centric web interfaces such as Windows Live Maps (also known as Windows Live Local, based on the Microsoft Virtual Earth technology at http://local.live.com ) and Google Maps (see http://maps.google.com ) are useful for visualizing spatially and geographically related data such as locations, weather, traffic, eco-environment etc. After Google Maps and Microsoft Virtual Earth published useful APIs to overlay location data on their maps, it is possible for customers to overlay their own data on top of browsable maps. This project will investigate how to publish, query, and visualize real-time data from live sensors over a geo-centric web interface. The main application area is environmental monitoring for eResearch.
Content-based 3D Object Retrieval
Lead Investigator Dr Jinglan Zhang
The number of 3D models available on the web is growing increasingly due to inexpensive hardware such as scanners and affordable interactive tools like CAD. 3D searching tools are needed to help people find these 3D models. Shape searching can be text-based using metadata such as keywords, file names, captions and context descriptions. However, the text (or symbol) -based matching cannot be used if the model is not annotated, the annotation is un-specific, the annotation is derivative, keywords are too common or keywords are unknown to users. In those cases, content-based shape retrieval is preferred. This research aims to develop automatic tools for shape-based retrieval. Colours and textures are not considered.
Indexing and Summarization of Multimedia Sensors Data
Leader Investigator: Dr. Dian Tjondronegoro
Image, sound, and video data can be collected constantly by sensors to study the environmental behaviour during a time period. As a result, the rapidly growing multimedia data needs to be indexed and summarized to assist analysis and retrieval. This project aims to: 1) Investigate tools and techniques to process sound/image data and store them in the appropriate form for indexing and summarization, 2) Applying data mining techniques to multimedia data and semantically identify key events.
Research Area: Business Services
Identifying the Requirements for a State-of-the-Art Business Process Simulation Framework
Lead Investigator: Dr. Moe Thandar Wynn.
Business process simulation is a relatively new area of study, whereby, the existing techniques used in process simulation for industries such as defence and manufacturing are adapted to analyse generic business process models. Business process simulation involves developing an accurate simulation model which reflects the behaviour of a process, including the data and resource perspectives, and then performing simulation experiments to better understand the effects of running that process. Business process simulation is regarded as an invaluable tool for process modelling due to its ability to perform quantitative modelling (e.g., cost-benefit analysis and feasibility of alternative designs) as well as stochastic modelling (e.g., external factors and sensitivity analysis).
A number of business process simulation tools already exist. However, existing tools mostly rely on the designer to provide information regarding processes and data during the development of a simulation model. Hence, the development of a simulation model is time consuming. Most do not have a direct link to the backend process execution engines or workflow systems. As a result, errors can be easily introduced in a simulation model. Furthermore, the existing tools also provide only limited support for the resource perspective.
The vision of this research is to build the foundation for a state-of-the-art and general business process simulation framework. Together with researchers from University of Eindhoven , a number of researchers at the Business Process Management Cluster at QUT are currently working on identifying the requirements of a business process simulation framework that supports rich resource behaviour and provides assistance in input modelling from workflow system logs.
We are interested in students with an overall GPA of 5.5 and above with research interest in the areas of business process modelling, process simulation, and process mining. We can tailor a project to suit a student's knowledge and interest.
Research Area: Information Science
Understanding and Promoting Health Information Literacy: Ageing Well in the Australian Community
Lead Investigator: Ms. Helen Partridge
This project aims to enhance peoples' health information literacy (HIL): their ability to take responsibility for their health by knowing how to look for, find and use health information. Ageing well is vital for Australia 's future. HIL will help Australians age well, to live longer with more years of good health, contributing actively to society and the economy. This project will develop the first model of health information literacy. It will make it possible for Australian libraries and other information agencies, to create information services to enhance the health information literacy of all Australians, and in particular our ageing population. This project will place Australia at the forefront of both consumer health information and information literacy research.
By 2051, 1 in 4 Australians will be aged 65 years and over; an extra $17.07 billion will be required to maintain health services to our ageing population. This project will develop the world's first model of health information literacy, an important foundation for creating a health smart society. In the health smart society Australians know how to seek, find and use information to help them age well. Understanding how people use information to age well will enable libraries to create services that will enhance the health information literacy of all Australians. This will reduce cost and other pressures on health and aged care systems, releasing funds to be directed towards other national health issues.
Project Objectives
This research aims to enhance and promote HIL in the Australian community. This will be achieved through an investigation of the experiences of HIL amongst Australians and through the development of the first model of HIL in community. The project has the following four key goals:
Goal 1: Discover and model the experience of HIL amongst Australians.
Goal 2: Extend the use of an innovative approach (i.e. phenomenography) to CHI research.
Goal 3: Establish a professional framework, based upon the HIL model developed.
Goal 4: Enhance the knowledge of those involved in supporting CHI about HIL and highlight opportunities for using the professional framework in the provision of quality CHI services, resources and training.
Granule Mining in Databases and its Applications in Share Trading
Lead Investigator: A/Prof. Yuefeng Li
Granule mining is a new technique that was proposed by QUT researchers recently. This technique can be used to improve the performance of data mining; and can be used in many application areas; for example, it can be used to predict if certain shares can get positive gains in a period of time based on current trading information of other shares.
Web Mining for Web Personalized Information Gathering
Lead Investigator: A/Prof. Yuefeng Li
This research discusses how to gather interesting and useful information from the Web to meet what users want. It focuses on how to find interesting patterns or ontology from Web data in order to describe Web user profiles efficiently. It also discusses techniques of using user profiles to quickly filter out noise or non-relevant information.
Recommender Systems
Leader Investigator: Dr. Yue Xu
The amount of information available on the Web has been increasing dramatically. Users are often overwhelmed by the huge amount of information and are faced with the great challenge of finding the most relevant piece(s) of information in a short amount of time. Significant research endeavours are being invested into building support tools that ensure the right information is delivered to the user quickly and accurately. Recommender systems are one of the tools designed to help users deal with the information explosion by giving information recommendations according to their information needs.
This project is to develop novel recommendation making algorithms that aim at solving two problems. One of the obstacles for building high quality recommender systems is the so called Cold Start problem which refers to the situation where a system does not have sufficient initial data resources to generate quality recommendations. Another problem is called recommendation novelty that refers to the fact that in many situations users prefer recommendations with serendipity rather than accuracy.
Concise Association Rule Mining
Leader Investigator: Dr. Yue Xu
For most of the work done in discovering association rules, the primary focus has been on the efficiency of the approach and to a lesser extent the quality of the derived rules has been emphasised. One big concern with the quality of association rule mining is the huge size of the extracted rule set. As a matter of fact, very often tens of thousands of association rules are extracted among which many are redundant thus useless in practice. The extremely large number of rules makes it difficult for the end users to effectively use the discovered rules. If the extracted knowledge can’t be effectively used in solving real world problems, the effort of extracting the knowledge is worth little. This is a serious problem but not yet solved satisfactorily. This project aims at investigating concise representations for association rules and novel data mining methods to discover non-redundant association rules.
Cross-language Information Retrieval
Leader Investigator: Dr. Yue Xu
The increasing diversity of the Internet has created a tremendous number of multilingual resources over the Web. A huge amount of Web documents are written in various languages other than English. Consequently, the demand for searching across language boundaries is growing exponentially. It is desirable that a search engine can deal with queries in different languages and search for information over collections of documents in multiple languages. This project aims at investigate techniques to improve the quality of query translation and result merging which are two key issues for multi-lingual information retrieval.
Delivering Interactive TV on the Web 2.0
Leader Investigator: Dr. Dian Tjondronegoro
The growing popularity of YouTube proves that users increasingly consume and share videos on the Web 2.0. Traditional TV broadcasters are starting to follow the phenomenon by delivering interactive TV on a set-top box which is connected to Broadband Internet to enrich the contents. This project aims to design new framework, toolsets and techniques to manage digital video contents, while exploring the opportunities for tagging, delivering and sharing of favorite videos on blogs, Podcasts, and Wikis. There will be a special focus on entertainment domain such as sports and news to encourage users to form virtual communities and enrich the interactions.
User-Centred Multimedia Web Search
Leader Investigator: Dr. Dian Tjondronegoro
Information providers are increasingly storing more image, sound, and video content on the Web. However, most of the current Web search engines still mainly support text-based retrieval which is not always useful for day-to-day requirements. This project aims to: 1) Study and model users' behaviour, tasks, and activity of multimedia web search, 2) Design, develop, and test a prototype that meets the trends and emerging requirements for next-generation Multimedia web search (based on the user model).
Research Area: Systems Science
Asset Management Decision Support Using Probabilistic Model Checking
Lead Investigator: Prof. Colin Fidge
Maintaining large-scale engineering assets, such as power stations and water supply systems, involves making renewal and operations & maintenance decisions with significant financial risks. A decision support environment is needed to help make such decisions, based on available data about the health and service history of the assets.This project will show how model checking can be used to aid the decision making process, using probabilistic data derived from past maintenance histories. To do this, formal models will be developed in a probabilistic model checker, such as PRISM, and will be used to simulate future maintenance requirements and costs.
An Integrated Asset Management Decision Support Tool
Lead Investigator: Prof. Colin Fidge
Maintaining large-scale engineering assets, such as power stations and water supply systems, involves making "renewal" and "operations & maintenance" decisions with significant financial risks. Making such decisions is challenging for asset managers in several ways: the decisions need to be based on multiple criteria; the outcomes need to satisfy multiple organisational and technical goals; and decisions need to be made at multiple time scales, ranging from long-term strategic planning to short-term operational actions. Although there are many separate tools and techniques to help with decision making - for example, decision trees, the Analytic Hierarchy Process, and fuzzy logic - what is lacking is a way of integrating their disparate and possibly contradictory recommendations. This project will develop a "plug in" software environment which allows different tools to be integrated smoothly and transparently, thus providing asset managers with well-founded and accountable support for making critical decisions.
Change Impact Analyses for Critical Embedded Programs
Lead Investigator: Prof. Colin Fidge
Embedded computer software controls a wide variety of security-critical and safety-critical systems, such as Supervisory Control and Data Acquisition (SCADA) systems. Maintaining long-lived computer-controlled systems is a challenging and costly task because the overall system typically outlives many generations of computer processors and peripheral devices. Ideally, hardware upgrades should not require us to completely redevelop trusted computer software, but only those parts that are dependent on hardware interaction.
Conducting a Change Impact Analysis is a standard, but largely informal, procedure for trying to isolate the effects of proposed hardware changes on critical software. In this project we will formalise the principles underlying impact analyses by applying program slicing, program refactoring and other rigorous and automatable techniques to the problem of tracing the influence of hardware changes. Apart from program code, the project will also consider the impacts of hardware changes on other programming artifacts such as requirements specifications and test suites.
Verified Spreadsheets for Mission-Critical Business Applications
Lead Investigator: Prof. Colin Fidge
Spreadsheet programs, such as Microsoft's Excel, have become an integral tool for collecting and managing data in most businesses.Spreadsheet software has become increasingly sophisticated and can now perform computational tasks that were formerly done by stand-alone applications programs. However, the users and developers of spreadsheets do not normally have a programming background, so the spreadsheets they produce can have subtle calculation or logic errors.
There are many well-documented cases of corporate collapses occurring due to spreadsheet errors, so there is a strong imperative to develop spreadsheets to the same standard of rigour as high-integrity computer programs. This project will determine how computer science principles normally used for verifying program code can be applied to spreadsheet formulae and macros. In particular, ways for formally proving spreadsheet calculations semantically correct will be developed, using the principles of pre- and post-conditions, and loop invariants and variants. Information flow principles will also be used to trace the dependencies between spreadsheet cells.
Dynamic Optimisation of Wireless Sensor Networks
Lead Investigator: Dr. Maolin Tang
A sensor is a small, inexpensive device with on-board sensing, processing/storage, transceiver and power units. A group of sensors communicating in a wireless medium for the purpose of gathering information and transmitting it to user nodes form a wireless sensor network. Wireless sensor networks have been applied in many areas, such as traffic control and military surveillance. This project investigates wireless sensor networks optimisation. The focus of this project is placed on dynamical topology discovery and dynamic routing under simultaneously considering energy efficiency, fault tolerance and computation/communication trade offs.
Government Web Services Management System
Lead Investigator: Dr. Maolin Tang
This project aims at improving the efficiency, quality, and productivity of Government's online services by building a government Web services management system. Web services are rapidly emerging as a popular standard for implementing online services. There will be more and more Government departments, organisations, and agencies moving towards a service-oriented architecture and using Web services to deliver online services. However, there is no Government Web Service Management System managing the Web services. The goal of the project is to develop a Government Web service management system which can: 1) discover and dynamically update available Government Web services; 2) support parallel access to multiple Web services; 3) build new Web services using existing Web services; 4) reconfigure composite Web services; 5) optimise the communications between the Web services; 6) plan and optimise the deployment of Government Web services.
Evolutionary Cutting, Packing and Layout
Lead Investigator: Dr. Maolin Tang
Cutting, packing and layout arise in a wide range of industries. The problem, typically, involves the placement of objects (possibly for subsequent cutting) onto sheets or rolls of material with the requirement that the objects must not overlap and must not exceed the boundaries of the sheets/rolls. This problem occurs in the paper, textile, leather, wood, glass and metal cutting industries, as well as many others. In general, the objective is to minimise waste material with constraints of different kinds. This has led to an increase in scientific interest in the automation of the packing process and the development of automation strategies. Since the problem's infancy, automated approaches have utilised many different techniques which include: linear and dynamic programming, problem specific heuristic methods, local search methods and, more recently, meta-heuristic methods such as genetic and evolutionary algorithms, artificial neural networks and simulated annealing. This project will focus on evolutionary algorithms.
Post processing of video sequences for the detection of roundabouts and road intersections with a road map
Lead Investigator: Dr Frederic Maire
The aim of the project is to detect the segments in a video sequence that correspond to roundabouts and road intersections. The method suggested would involve fusing information from a road map with features detected in the images.
Post processing of video sequences for the localization of road signs
Lead Investigator: Dr Frederic Maire
The aim of the project is the development of a system for road sign detection based on edge orientation histograms and colours. Edge orientation histograms are reliable, scale and contrast invariant features that can be extracted efficiently using integral images. Colour information will be used to make the system more robust.
Post processing of video sequences for the detection of pedestrians
Lead Investigator: Dr Frederic Maire
The aim of the project is the development of a system for pedestrian detection based on face detection. The method suggested would involve using a standard face detection algorithm (like Viola Jones algorithm) to detect people at close range and backtrack in the videos to annotate the corresponding blobs earlier in the video sequence.
Research Collaborations
Information Security Institute
File Carving Fragmented Hard Drives
Lead Investigator: Dr. Andrew Clark
File carving is the process of recovering files, including fragmented files, from disk images where the directory data has been lost. File carving is particularly useful for computer forensic investigators who may be wish to extract potential evidence from damaged or deleted regions of hard disk drives. Current file carving techniques consist of two main tasks: identifying the blocks which potentially correspond to a complete file; and then using a validation step to see if the file appears to have been correctly reconstructed (based upon the file type and its encoding).
Recovering fragmented files is particularly difficult as regions of a single file may be distributed widely across the disk media. This project will investigate techniques for reconstructing fragmented files using approaches such as combinatorial optimisation. The project will involve building upon existing libraries of known file types, and developing an appropriate representation for such information, including information about the nature of the file content. Additionally the project will experiment with new approaches for selecting fragmented regions of the media corresponding to single files.
Volatile Memory Forensics
Lead Investigator: Dr. Andrew Clark
Acquisition and analysis of volatile memory from computing devices are difficult problems. This project will focus on the analysis of acquired volatile memory dumps with a focus on identifying "useful" information from within the dump. Such information includes passwords, cryptographic keys, cleartext (pre-images) or encrypted data.
Initially this project will focus on case studies of specific software tools in order to better understand how they store sensitive information within volatile memory, and whether or not simple, application-specific algorithms can be utilised for extracting that information. An outcome of this phase of the work will be a classification scheme which maps the different types of sensitive information to how they are usually stored in memory. Following on from this initial step, the project will aim to develop general techniques for identifying and extracting the various classifications of sensitive information from captured volatile memory.
Change Impact Analyses for Critical Embedded Programs
Lead Investigator: Prof. Colin Fidge
Embedded computer software controls a wide variety of security-critical and safety-critical systems, such as Supervisory Control and Data Acquisition (SCADA) systems. Maintaining long-lived computer-controlled systems is a challenging and costly task because the overall system typically outlives many generations of computer processors and peripheral devices. Ideally, hardware upgrades should not require us to completely redevelop trusted computer software, but only those parts that are dependent on hardware interaction. Conducting a Change Impact Analysis is a standard, but largely informal procedure for trying to isolate the effects of proposed hardware changes on critical software. In this project we will formalise the principles underlying impact analyses by applying program slicing, program refactoring and other rigorous and automatable techniques to the problem of tracing the influence of hardware changes. Apart from program code, the project will also consider the impacts of hardware changes on other programming artefacts such as requirements specifications and test suites.
Top
Security for Wireless Sensor Networks
Lead Investigator: Dr. Ernest Foo
Wireless sensor networks consist of small self-powered wireless devices that communicate to a base station. Sensor networks are often used for secure applications but the power and communication restrictions have limited the security available. Recently TPM chips have been incorporated in these devices and libraries have been written to allow them to conduct bilinear pairings. This project will investigate the use of bilinear pairings in wireless sensors to develop new efficient mechanisms for key distribution, authentication, data aggregation and confidentiality.
Seamless and secure mobile communications
Lead Investigator: Dr. Jason Smith
Mobile computing platforms, including laptop computers, Internet tablets, and mobile smartphones now commonly integrate wireless chipsets that permit these devices to access network resources and services in personal, local, metropolitan, and wide area contexts.As such connectivity becomes ubiquitous two important challenges must be addressed: (1) techniques for securing communications end-to-end across a range of layer two technologies must be developed, and (2) the signalling protocols used to manage mobility and the migration from one communications layer to another must also be secured.This project will investigate the security requirements of emerging mobile networks that can utilise a multitude of layer-two technologies and identify or develop protocols suitable for ensuring the security of both user and signalling data in such networks.
Command and Control in Distributed Denial of Service Attacks
Lead Investigator: Dr. Jason Smith
Distributed denial of service (DDoS) attacks remain a persistent feature of, and significant threat to, the Internet. While a number of approaches have been proposed to improve the effectiveness of responses to DDoS attacks, a more promising direction is the disruption of the command and control channels that attackers must utilise to direct attacks. Traditionally DDoS attackers have utilised centralised command and control channels (internet relay chat rooms for example), but in recognition that such centralised command and control approaches are a point of vulnerability, they have started to utilise more distributed, peer-to-peer based methods of control. This project will investigate the evolution of command and control in DDoS attacks and develop techniques to aid in the disruption of DDoS when distributed command and control channels are employed.
Privacy Preserving Computational Grids
Lead Investigator: Dr. Jason Smith
Grid or utility computing enables the processing power of distributed computing resources to be harnessed on demand. This capability has been widely adopted by the scientific community, but if such a paradigm is to be adopted more generally by business and industry, grid computing architectures must be able to ensure the confidentiality of both the data and the algorithms used to process that data. This project will investigate grid computing techniques capable of protecting commercially sensitive data and algorithms that will encourage the utilisation of computational grids by business and industry.
Design of Secure Mobile GIS System
Lead Investigator: Ms. Rong Du
Mobile devices will be widely used in GIS field trips and various data accessing applications. The devices use different types of communication protocols not only to access public data but also restricted data sets and services, such as sensitive information or services through payment. Using mobile devices in various GIS applications will generate unknown security risks. This project will examine the use of mobile devices to access GIS data sets and services, the associated communication protocols, and their interaction with authentication and authorization services. The research outcomes will include secure protocol solutions verified using formal methods, threat models and test data generation for implementation, and protocol performance analysis. Prototype implementation may also be part of the project. Efficient cryptographic schemes for lightweight devices, such as pairing base encryption and digital signature schemes, may also be studied.
Top
Security Proofs with Automated Assistance
Lead Investigator: Prof. Colin Boyd
Reductionist proofs are a standard technique for gaining confidence in the security of cryptographic systems. Such proofs have a reputation for being difficult to check and easy to get wrong. Automated techniques for checking security of systems has a long history in the situation where cryptography is treated as a black box (the Dolev-Yao model) but it is only recently that machine support for reductionist proofs has been considered. This project will review current research in this direction, identify areas where improvement is possible, and apply chosen tools to derive and/or check proofs for a variety of cryptographic primitives and protocols.
Applications of Composite Order Bilinear Pairings in Cryptography
Lead Investigator: Juan Gonzalez Nieto
The recent discovery of the useful properties of bilinear pairings on elliptic curves has opened up a whole new area of cryptography based on simplified public key infrastructures. Many new protocols in this area have been proposed, mostly using pairings defined on groups of prime order. Bilinear pairings can also be defined over composite order groups and recent research has shown their potential in cryptography. This project will investigate applications of composite order pairings in cryptography. Expected outcomes include new cryptographic protocols for secure e-commerce.
Authentication for Small RF Devices
Lead Investigator: Dr. Juan Gonzalez Nieto
The next phase of the computer revolution will see ubiquitous use of small wireless computing devices. Prominent examples are radio frequency identification (RFID) tags and wireless sensor networks. The combination of mobility, wireless communications and low-power hardware presents unique challenges in the design of authentication protocols suitable for ubiquitous computing. This project will investigate location privacy and proximity authentication. Location privacy is needed to avoid illegitimate tracking of mobile devices. Proximity authentication is indispensable to avoid relay attacks. The anticipated outcomes will include new protocols and formal models in which to analyse their security properties.
Privacy Protecting Access Control
Lead Investigator: Dr. Jason Reid
This project will investigate how authorisation policy languages such as XACML can be used to automatically enforce privacy constraints over usage and dissemination of personal information in federated systems based on Web services standards. To comply with privacy legislation, organizations that collect personal information must ensure that subsequent uses and disclosures of the information are consistent with the purpose notified to the individual when the information was collected. Since the amount of personal information that is stored, processed and shared electronically is rapidly increasing, the task of ensuring that data is handled in a manner consistent with the disclosed purpose is becoming ever more difficult. In order to automate this process, systems need to be developed to tag personal information with privacy relevant metadata (disclosed purpose of collection, retention period etc.) so that access control systems can evaluate requests to determine whether they are consistent with privacy constraints. This project will also investigate methods for determining the likely purpose of a user who is requesting access to personal information based on contextual information such as job function, past access patterns, workflow progress etc.
High Assurance Information Sharing Networks
Lead Investigator: Dr. Jason Reid
There is a growing need for secure communication systems based on internet technologies to support information sharing between organizations. For example, infrastructure operators need to share highly sensitive information (vulnerability reports, incident reports etc.) to support critical infrastructure protection activities in the telecommunications, finance and energy sectors. Sensitive information requires controlled distribution and ongoing control over dissemination and usage - a form of Digital Rights Management. One of the key challenges in dissemination control is establishing trust in client platforms that connect to the information sharing network. This project will investigate a combination of trusted computing, secure operating systems, hardware virtualisation, cryptographic access control and a peer-to-peer network underlay for storage and distribution to provide a high-assurance information sharing environment.
Top
Stream Cipher Initialisation Schemes
Lead Investigator: Dr. Leonie Simpson
Recent design criteria for secure stream ciphers include an internal state size which is at least twice the key size, to prevent time-memory-data tradeoff attacks. Thus an initialisation scheme which expands the key to form the initial internal state is required. The initialisation process frequently includes the use of a known initialisation value such as a frame or packet number, so that new initial states can be formed from the same secret key. This is useful in maintaining synchronisation for applications such as mobile telephony or internet communications. This project aims to establish design criteria for initialisation schemes, so that the secret key remains protected, even if multiple initial states are known.
Authenticated Encryption
Lead Investigator: Dr. Leonie Simpson
Authenticated encryption schemes aim to use a symmetric encryption scheme to provide both confidentiality and authentication for data communications. Six of the thirty four stream ciphers submitted to the current European eSTREAM project claimed to provide authenticated encryption. All were shown to be insecure. An investigation of these, and possibly other, proposals will be performed to identify structural features of the stream cipher designs which contribute to their vulnerability, and devise design criteria for authenticated encryption schemes which offer greater security against known or chosen plaintext attacks.
Effective Information Security Assurance Management in Contemporary Organizations
Lead Investigator: Dr. Lauren May
In the Information Age, contemporary organisations are heavily reliant on their information and information infrastructures as foundations of their business. The majority of these organisations make use of Internet technologies as a means of fast, efficient and economical communications in their daily business operations. In recent years the management of information security has become a topical focus due to a number of factors at both the governance and operational levels.This research proposes to investigate particular aspects* of the processes for attaining effective information security management in contemporary organizations in the Information Age.
* Note that these particular aspects depend upon the interests, background and skill set of the research student. They may be at a management level, a technical level, or any combination. The research may also be industry-linked; for example linked to the student's work environment.
Cooperative Research Centre – Smart Services
One-Stop Personalised Financial Services
Lead Investigator: See below
Are you interested in understanding how people use technology, or designing technology to meet peoples needs? Working with the One Stop Shop Project might be for you. Projects are available in the following areas:
- User studies into how people think about their personal finances and the key questions they wish to answer (with respect to basic banking and payment services). Key contact: Associate Professor Judy Drennan
- The application of Web 2.0 and other social networking techniques to financial services, making them more attractive to “Gen Y”. Key contact: Professor Kerry Raymond
- User studies into customers' trust and privacy issues relating to financial institutions and their online services, especially the privacy implications of merging previously separate data. Key contact: Associate Professor Audun Josang
- Data mining to help populate a whole of life financial information model. Key contact: Dr Richi Nayak
One Stop Shop is a project of the Smart Services Cooperative Research Centre. It brings together financial institutions and researchers interested in people's use of technology and information, technical architecture and data mining.
More about the project
Currently financial services are provided to the customer through a set of disconnected interfaces and the individual services are presented in a provider-centric (a chronologic list of payments to and from accounts) rather than a customer-centric manner. For example, current systems do not assist people in answering questions such as:
- how much is my car costing me each year?
- will I have enough money to pay the electricity bill next month?
- can I afford a $5000 holiday next January?
- how much should my partner and I save each week to send our kids to private schools?
- when can I afford to retire?
Objectives
This project aims to:
- understand how customers think about their personal finances and their financial literacy
- develop online and mobile interfaces that reflect the customer's point of view about their finances, including controls to allow customers to manage their own privacy and security profiles
- develop technical architectures to underpin a “plug and play” service that reflects the user's service needs
- understand the customer's trust and privacy issues to deliver financial education in new-to-the-customer financial services and the online and mobile interfaces to those services
Building and Leveraging User Profiles for Personalisation and Search
Lead Investigator: Dr Richi Nayak and Professor Amanda Spink
This large project takes a user-centric approach to anticipating and offering services that customers are predicted to want in the near future. This project will explore and develop techniques for acquiring knowledge about individual service customers. User profiles and market segmentation models are constructed capturing data on preferences of service customers, customer behaviour, and social networks of customers. Techniques are explored to leverage user profiles for searching and personalisation of contents by selecting relevant content and designing service presentation in real time.
More specifically, this project will develop and apply data mining and knowledge acquisition techniques (such as clustering, classification, association mining, and ontology mining) for the construction of user profiles and/or market segmentation models based on available data (such as Web pages, Web access logs, registration data, customer profile portfolios, etc.). This project will further develop and apply new data mining and knowledge acquisition techniques for the construction of information filtering models and recommendation models utilising user profiles and market segmentation models, to be used in personalised services and frameworks.
Trends in User-Generated Content, Media User Behaviour, Mobility And Use Of Social And Online Media
Lead Investigator: Professor Amanda Spink
This large project examines how audience use of media is likely to change with a specific focus on the youth segment (under 25’s) and baby boomers (45+). The project explores the underlying drivers and trends such as user-generated content, lead-user behaviour, mobility and increasing use of social media will be incorporated into findings. Lead-user behaviour will be identified through a media use based survey of the literature and social probes on lead users, focusing on under 25’s and over 45’s.
A large-scale state/national level survey baseline study will be conducted to examine how Australians' are using online services and the impact of these services on Australian life and the economy. Audience knowledge discovery tools will be developed to mine customer data across a media organisation for the purposes of trend detection. The study will examine: (1) current user patterns and experience in interacting with existing services based on quantitative data provided by media participants, and (2) current and emerging user patterns and experience in interacting with online information and entertainment services.
A number of PhD projects are available in the areas of “Trends in user-generated content, media user behaviour, mobility and use of social and online media”.
Cross-Media Delivery of Adaptive and Pervasive Services
Lead Investigator: Dr Dian Tjondronegoro
This large project will examine and develop techniques for the delivery of services which are customised according to awareness of the service user’s delivery channel and platform. The same mechanisms may also be used to re-purpose multimodal contents using awareness of the service user’s identity and preferences. There will be a special focus on personalisation of services for emerging mobile platforms and networks, such as wireless broadband network, and on in-vehicle environments.
More specifically, this project will survey, develop and apply techniques for repurposing cross-media contents to suit current and next generation devices, contexts, and user profiles. This project aims to design a framework to enable consistent experience across on-line and small mobile devices, taking into account emerging hand-held device technologies and adaptive interfaces.
A number of PhD scholarships are available in the areas of:
- “Automatic processing and content repurposing of cross-media contents (video, image, and text) for pervasive and personalised services”
- “Adaptive interaction design for online and small mobile devices over wireless broadband network”