Since May 2011, we started to organise Computing HDR seminars. Each invited speaker is a Computing HDR student who has received a Best Student Paper Award or a Best Paper Award at a reputable conference, or has a publication in an ERA/CORE rank A* or leading conference/journal. The aim to organise HDR seminars is to introduce some Computing HDR students and their successful research experience and story to the others. By this means, we expect to see more high quality research work from Computing HDR students.
Who to Contact?
If you have good research work to introduce at an HDR seminar, please contact A/Prof. Yan Wang.
Upcoming Seminars - To Be Advised
Previously Organised Seminars
- Semiars on 17 May 2016
- Seminar on 22 March 2016
- Seminar on 11 June 2015
- Seminar on 15 November, 2013
- Seminar on 22 November, 2012
- Seminar on 12 November, 2012
- Seminar on 1 November, 2012
- Seminar on 2 June, 2011
- Seminar on 19 May, 2011
- Seminar on 5 May, 2011
Seminar on 17 May, 2016 in Building E6A Rm357
Time: 2:00 PM – 3:00 PM
Time: 3:00 PM – 4:00 PM
Speaker: Robertus Nugroho
Title: Time-sensitive Topic Derivation in Twitter
Much research has been concerned with deriving topics from Twitter and applying the outcomes in a variety of real life applications such as emergency management, business advertisements and corporate/government
Robertus Nugroho is currently a PhD Student in the Department of Computing, Macquarie University, under the supervision of Prof. Jian Yang. He is also awarded a postgraduate studentship at CSIRO Data61 from 2014 to 2017. Robertus received a master degree in computing and information technology from The University of New South Wales in 2009. Recently, he received a Digital Productivity Staff Award from CSIRO Data61 for the significant contribution in support of scientific outcomes (December 2015). His current research interests include social media and big data analytics, machine learning, and web technology.
Award: Best Paper Award
R. Nugroho, W. Zhao, J. Yang, C. Paris, S. Nepal, Y. Mei: Time-sensitive topic derivation in twitter. In: Web Information Systems Engineering – WISE 2015: 16th International Conference, Miami, FL, USA, November 1-3,
Speaker: Farshid Anvari
Title: Effectiveness of Persona with Personality on Conceptual Design
User Centred Design (UCD) is a methodology used to develop applications that consider the goals of the users as a primary requirement. The use of personas, archetypical users, in UCD is well established in the software industry. Personas are used to facilitate the design of applications focusing on target users and to communicate with stakeholders. Personas may take various forms: personas, mash-up personas, incomplete personas and unspoken personas.
Holistic Personas, a persona enriched with personality traits, is deemed to better represent a user as personality provides a richer profile and affects the way users interact with technology. Our study explores the effects of the use of Holistic Personas on participants’ performance in creating conceptual designs.
Four Holistic Personas with different personalities were presented to 91 participants, in studies in Australia and Denmark and collectively they completed 218 design artefacts.
Our results indicate that the participants were able to identify the personality traits of the Holistic Personas and their ratings of the personalities matched closely with the intended personalities. The findings support that personas with personality traits can aid software engineers to produce requirements and conceptual designs tailored to the needs of specific personalities.
Personnel engaged in developing applications using UCD techniques need to have special abilities and training to design products that meet the needs of users. Thirty-three participants completed a spatial ability test, answered personality trait questionnaires and performed a design activity using one of the Holistic Personas. Our assessment of design artefacts indicate that participants who score high in imagination personality factor and spatial ability tests are talented designers in the use of personas with personality within UCD methodologies. The implication of our study is that the talented designers can be identified and utilised more productively.
Farshid Anvari holds a Master of Philosophy (MPhil), Macquarie University (2016), a Graduate Diploma of Information Technology, University of Southern Queensland (1998), a Graduate Diploma of Science, University of Tasmania (1996), and a Bachelor of Engineering, Swinburne University of Technology (1984). His MPhil supervisors were Prof. Deborah Richards and A/Prof. Michael Hitchens.
He has over 15 years of experience in software engineering. He was working at Australian Biosecurity Intelligence Network (ABIN) CSIRO, University of New South Wales (UNSW) and Special Broadcasting Service (SBS). At ABIN he developed an application for scientists to disseminate tools. At UNSW he architected solutions and led the development team for implementation of Healthy Me: an online multi-research application, supporting consumers in informed decision making about their health. At SBS he designed and implemented the automated digital channel system, SBS Essential, which was highly commended in the Commonwealth Broadcasting Association awards for cost-effective engineering. He also has over 12 years of experience in power engineering, specialising in hydro system modelling and reliability studies. His research interests include cognitive load, user centred design, persona ontology and architecting systems that are economical, robust and reliable.
Mr. Anvari is a member of the Australian Computer Society and the Association for Computing Machinery.
Anvari, F., Richards, D., Hitchens, M. and Babar, M. A. (2015) Effectiveness of Persona with Personality Traits on Conceptual Design. In Proceedings of the ICSE 37th International Conference on Software Engineering (CORE2014 rank A*), Florence, Italy, May 2015. pp. 263-272.
Seminar on 22 March, 2016
Speaker: Kinzang Chhogyal
Title: Probabilistic Belief Contraction Using Argumentation
The epistemic or belief state of an intelligent agent is generally in a state of flux with new beliefs being added and old beliefs being removed. The field of belief change deals with the modelling of different types of changes an agent’s belief state may undergo. Contraction is one of the major belief change operations where a sentence loses its status of a belief and is demoted to a non-belief, i.e. it is neither believed nor disbelieved. When a belief state is represented as a probability function P, the resulting belief state after contracting a belief h from the original belief state P can be given by the mixture of two states: the original state P, and the resultant state P' of revising P by the negation of h. However, it is not clear exactly what these proportions should be and there is no work in literature addressing this. The primary aim of this work is to propose a plausible solution to this problem. We begin by first classifying different belief states according to their stability, and then exploit the quantitative nature of probabilities and combine it with the basic ideas of argumentation theory to determine the mixture proportions. We, therefore, propose a novel approach to the problem of probabilistic belief contraction using argumentation.
Kinzang Chhogyal recently submitted his PhD thesis in Computer Science/Artificial Intelligence to Griffith University in Brisbane. Much of the work in his dissertation was undertaken while he was a visiting student at Macquarie University and where he worked under the supervision of Assoc. Prof. Abhaya Nayak. His current research interests include knowledge representation and reasoning, belief change and cognitive science.
Chhogyal, Kinzang, A. C. Nayak, Zhiqiang Zhuang, and Abdul Sattar. "Probabilistic belief contraction using argumentation." In Proceedings of the Twenty-Fourth International Joint Conference on Artificial Intelligence, IJCAI, pp. 25-31. 2015.
Seminar on 11 June, 2015
Speaker: Xiaoming Zheng
Title: Trust Prediction with Propagation and Similarity Regularization
Online social networks have been used for a variety of rich activities in recent years, such as investigating potential employees and seeking recommendations of high quality services and service providers. In such activities, trust is one of the most critical factors for the decision-making of users. In the literature, the state-of-the-art trust prediction approaches focus on either dispositional trust tendency and propagated trust of the pair-wise trust relationships along a path or the similarity of trust rating values. However, there are other influential factors that should be taken into account, such as the similarity of the trust rating distributions. In addition, tendency, propagated trust and similarity are of different types, as either personal properties or interpersonal properties. But the difference has been neglected in existing models. Therefore, in trust prediction, it is necessary to take all the above factors into consideration in modelling, and process them separately and differently.
In this talk, we introduce a new trust prediction model based on trust decomposition and matrix factorization, considering all the above influential factors and differentiating both personal and interpersonal properties. In this model, we first decompose trust into trust tendency and tendency-reduced trust. Then, based on tendency-reduced trust ratings, matrix factorization with a regularization term is leveraged to predict the tendency-reduced values of missing trust ratings, incorporating both propagated trust and the similarity of users’ rating habits. In the end, the missing trust ratings are composed with predicted tendency-reduced values and trust tendency values. Experiments conducted on a real-world dataset illustrate significant improvement delivered by our approach in trust prediction accuracy over the state-of-the-art approaches.
Biography: Xiaoming Zheng is currently a PhD candidate in the Department of Computing, Macquarie University. He received his B.Eng degree in Automation from North University of China, P.R. China in 2008, and M.Eng degree in Control Theory and Control Engineering from Shanghai University, P.R. China in 2011. His research interests include trust prediction in online social networks and EEG pattern recognition in spontaneous BCI. At MQ, his supervisors are Prof. Mehmet A. Orgun and A/Prof. Yan Wang.
Publication: X. Zheng, Y. Wang, M. A. Orgun, Y. Zhong and G. Liu, Trust Prediction with Propagation and Similarity Regularization, AAAI2014, July 27-31, 2014, Quebec City, Quebec, Canada
Seminar on 15 November, 2013
Speaker: Byungho Min
Title: Antivirus security: naked during updates
The security of modern computer systems heavily depends on security tools, especially on antivirus software solutions. In the anti-malware research community, development of techniques for evading detection by antivirus software is an active research area. This has led to malware that can bypass or subvert antivirus software. The common strategies deployed include the use of obfuscated code and staged malware whose first instance (usually installer such as dropper and downloader) is not detected by the antivirus software. Increasingly, most of the modern malware are staged ones in order for them to be not detected by antivirus solutions at the early stage of intrusion. The installers then determine the method for further intrusion including antivirus bypassing techniques. Some malware target boot and/or shutdown time when antivirus software may be inactive so that they can perform their malicious activities. However, there can be another time frame where antivirus solutions may be inactive, namely, during the time of update. All antivirus software share a unique characteristic that they must be updated at a very high frequency to provide up-to-date protection of their system. In this paper, we suggest a novel attack vector that targets antivirus updates and show practical examples of how a system and antivirus software itself can be compromised during the update of antivirus software. Local privilege escalation using this vulnerability is also described. We have investigated this design vulnerability with several of the major antivirus software products such as Avira, AVG, McAfee, Microsoft, and Symantec and found that they are vulnerable to this new attack vector. The paper also discusses possible solutions that can be used to mitigate the attack in the existing versions of the antivirus software as well as in the future ones.
Biography: Byungho Min is currently a PhD candidate in the Department of Computing, Macquarie University. He received his B.Sc and M.Sc in Computer Science and Engineering from Seoul National University (SNU), Republic of Korea. He worked in the Ministry of Defence, Republic of Korea as an IT security expert for 4+ years. His research interests include Smart Grid and SCADA security, malware analysis and offensive techniques. His supervisors are Prof. Vijay Varadharajan, A/Prof Michael Hitchens and Dr Udaya Tupakula.
Publication: Byungho Min, Vijay Varadharajan, Udaya Tupakula and Michael Hitchens: Antivirus security: naked during updates. Software: Practice and Experience. (2013, Accepted)
Seminar on 22 November, 2012
Speaker: Haibin Zhang
Title: Efficient Contextual Transaction Trust Computation in E‐Commerce Environments
In e-commerce environments, trust is a dominating factor in seller selection. Most existing trust evaluation studies compute a single trust value to reflect the “general” or “global” trust level of a seller provider without any information of transaction context taken into account. As a result, a buyer may be easily deceived by a malicious seller in a forthcoming transaction. For example, with the notorious “value imbalance problem”, i.e., a malicious seller can build up a high trust level by selling cheap products and then starts to deceive buyers in selling expensive products.
To detect this problem, which is actually one type of the transaction context imbalance problem, and avoid massive monetary losses of buyers, the trust results (referred to as Contextual Transaction Trust, CTT for short) should be transaction context based and have the capacity to outline the seller’s reputation profile indicating the trust level in a specific product or a product category, a price range, a time period or any necessary combination of them. However, it requires the pre-computation of aggregates over large-scale ratings and transaction data with necessary combinations of three dimensions (i.e. product category, price and time), so as to deliver real-time responses to a buyer’s query.
Towards solving this challenging problem, we propose three methods, i.e., aR-tree based approach, aP-tree based approach and a hybrid structure of aP+-tree and aB+-tree to fast indexing rating aggregates for CTT computation. The experimental results illustrate the performance of these methods in responding to CTT queries, respectively. To the best of our knowledge, this is the first solution in the literature to the computation of CTT in e-commerce environments.
Biography: Haibin Zhang is currently a PhD candidate in the Department of Computing, Macquarie University, Australia. He received his BEng degree in computer science and technology from the Harbin Engineering University (HEU), P.R. China in 2007, and MEng degree in computer science and technology from the Harbin Institute of Technology (HIT), P. R. China in 2010. His current research focuses on context-aware trust computation in e-commerce applications. At MQ, his supervisors are Assoc/Prof. Yan Wang and Prof. Mehmet A. Orgun.
Award: Best Paper Award
Publication: Haibin Zhang, Yan Wang and Xiuzhen Zhang, Efficient Contextual Transaction Trust Computation in E-Commerce Environments, 11th IEEE International Conference on Trust, Security and Privacy in Computing and Communications (TrustCom-2012) (acceptance rate: 100/358 = 28%), 25-27 June 2012, Liverpool, UK
Speaker: Lan Zhou
Title: Enforcing Role-based Access Control for Secure Data Storage in the Cloud
With the rapid developments occurring in cloud computing and services, there has been a growing trend to use the cloud for large-scale data storage. This has raised the important security issue of how to control and prevent unauthorised access to data stored in the cloud. One well-known access control model is the role-based access control (RBAC), which provides flexible controls and management by having two mappings, users to roles and roles to privileges on data objects. Controlling the access in cloud systems is different from traditional access control systems, where access control enforcement is carried out by a trusted authority which is usually the service provider. In a cloud system, as data can be stored in distributed data centres, there may not be a single central authority which controls all the data centres. We propose a role-base encryption (RBE) scheme which integrates the cryptographic techniques with RBAC. Our scheme allows the owner of data to store it in an encrypted form in the cloud and to grant access to that data for users with specific roles. The scheme specifies a set of roles to which the users are assigned, with each role having a set of permissions. The data owner can encrypt the data and store it in the cloud in such a way that only users with specific roles can decrypt the data. Anyone else, including the cloud providers themselves, will not be able to decrypt the data.
Biography: Lan Zhou is currently a PhD student in Department of Computing in Macquarie University. His research interests include role-based access control, secure cloud storage, secure credential systems, and trust management system. Lan Zhou obtained his Masters Research Degree in Computer Science at the University of Wollongong, and Bachelor Degree at Huazhong University of Science and Technology, China. His PhD supervisors are Prof. Vijay Varadharajan and Assoc/Prof. Michael Hitchen.
Award: The Computer Journal Wilkes Award for 2011
Publication: Lan Zhou, Vijay Varadharajan and Michael Hitchens: Enforcing Role-Based Access Control for Secure Data Storage in the Cloud.Computing. J. 54(10): 1675-1687 (2011)
Seminar on 12 November, 2012
Speaker: Dr Raghav Ramachandran
Title: Probabilistic Belief Contraction
Probabilities have often been used to represent a rational agent’s knowledge/belief state. Any sentence with probability 1 is considered to be a ‘belief of the agent’. Probabilistic belief contraction refers to the problem of reducing the probability of a sentence from 1 to strictly below 1. Probabilistic belief contraction has been a much neglected topic in the field of probabilistic reasoning. This is due to the difficulty in establishing a reasonable reversal of the effect of Bayesian conditionalization on a probabilistic distribution. The uncertainty associated with a probability function is measured in terms of entropy measures. The application of ‘measures of uncertainty’ or entropy measures to perform probabilistic belief contraction was suggested by E. Olsson which lead to an account of probabilistic belief contraction proposed by A. Ramer based on the principle of maximum entropy. The principle of maximum entropy chooses that probability distribution which assumes the least, among those that satisfy the necessary conditions. The account of probabilistic contraction proposed by Ramer, however, results in unnecessary and excessive changes to the initial probability distribution. This solution needs to be improved. We study this problem using both the Shannon entropy measure as well as the Hartley entropy measure, with an aim to avoid excessive loss of beliefs.
Biography: Dr. Raghav Ramachandran graduated with a PhD from the Department of Computing, Macquarie University in 2012. His PhD thesis titled ‘Three Studies in Belief Removal’ was supervised by Dr. Nayak and Prof. Orgun. He also has a Master degree in Mathematics from the Indian Institute of Technology Madras. He is currently a research associate at the Computing Department, Macquarie University. He has been a visiting research fellow at the School of Information and Communication Technology, Griffith University, Queensland. His research are focused in the areas of probabilistic reasoning, belief revision and description logics. His PhD supervisors were Dr. Anhaya Nayak and Prof. Mehmet A. Orgun.
Publication: Raghav Ramachandran, Arthur Ramer and Abhaya Nayak: Probabilistic Belief Contraction, Minds and Machines 22 (4), pp. 325-351, 2012. Springer
Seminar on 1 November, 2012
Speaker: Dr Hadi Mashinchi
Title: Verification of A Smart Incontinence Monitoring System and The Use of Machine Learning for Analysing Sensor Data
This paper proposes a method of analysing and using data collected from a sensor embedded in a continence pad which is a component in the SIM™ wireless incontinence monitoring system. In contrast to systems described in the literature, SIM™ does not only detect urinary events but estimates volume. An optimized volume estimation model is developed using data remotely collected by SIM™ through the application of a machine learning algorithm. For verification of the model, a new objective function is introduced based on an understanding of the requirements of caregivers in nursing homes. The emulation results reveal that SIM™ makes recommendations which can prevent a resident being kept in a soaked pad with a performance of greater than 90%.
Biography: Hadi Mashinchi received his PhD degree in computer science from Macquarie University, Sydney, Australia in 2012. He is currently working as data mining specialist in Sydney. His research interests are hybrid optimization, global continuous optimization, fuzzy regression, and the application of data mining in engineering applications, such as campaign analytics, sensor data and etc. He has organized a few special sessions and served on the review boards of journals and international conferences. His PhD supervisors were Prof. Mehmet A. Orgun and Dr. Anhaya Nayak.
Publication: M. H. Mashinchi, M. A. Orgun, M. Mashinchi, W. Pedrycz: A Tabu-Harmony Search-Based Approach to Fuzzy Linear Regression. IEEE Transactions on Fuzzy Systems 19(3): 432-448 (2011)
Joint Honours and HDR Seminar on 2 June, 2011
Speaker: Abeed Sarker
Title: Improved Reconstruction of Flutter Shutter Images for Motion Blur Reduction
Relative motion between a camera and its subject introduces motion blur in captured images. Reconstruction of unblurred images is ill-posed due to the loss of spatial high frequencies. The flutter shutter preserves high frequencies by rapidly opening and closing the shutter during exposure, providing greatly improved reconstruction. We address two open problems in the reconstruction of unblurred images from flutter shutter images. Firstly, we propose a noise reduction technique that reduces reconstruction noise while preserving image detail. Secondly, we propose a semi-automatic technique for estimating the Point Spread Function of the motion blur. Together these techniques provide substantial improvement in reconstruction of flutter shutter images.
Biography: Abeed Sarker is now a PhD candidate in the Department of Computing, Macquarie University (enrolled in April 2010). He received his Bachelor of Information Systems (HONS) degree from Macquarie University in 2009 . His Honours research focused on digital image processing under the supervision of Dr. Leonard Hamey. His current research focuses on automatic text summarization for the medical domain and his supervisors are Dr. Diego Molla-Aliod (principal supervisor) and Dr. Cecile Paris (co-supervisor).
Award: Best Paper Award
Publication: Abeed Sarker, Leonard G. C. Hamey, “Improved Reconstruction of Flutter Shutter Images for Motion Blur Reduction”, 2010 International Conference on Digital Image Computing: Techniques and Applications (DICTA, 37 papers were selected for oral presentation from 194 submissions), 2010, pp.417-422
Seminar on 19 May, 2011
Speaker: Guanfeng Liu
Title: A Heuristic Algorithm for Trust-Oriented Service Provider Selection in Complex Social Networks
In a service-oriented online social network consisting of service providers and consumers, a service consumer can search trustworthy service providers via the social network. This requires the evaluation of the trustworthiness of a service provider along a certain social trust path from the service consumer to the service provider. However, there are usually many social trust paths between participants in social networks. Thus, a challenging problem is which social trust path is the optimal one that can yield the most trustworthy evaluation result. In this paper, we first present a novel complex social network structure and a new concept, Quality of Trust (QoT). We then model the optimal social trust path selection with multiple end-to-end QoT constraints as a Multi-Constrained Optimal Path (MCOP) selection problem, which is NP-Complete. For solving this challenging problem, we propose an efficient heuristic algorithm, H-OSTP. The results of our experiments conducted on a large real dataset of online social networks illustrate that our proposed algorithm significantly outperforms existing approaches.
Biography: Guanfeng Liu is a now PhD candidate in the Department of Computing, Macquarie University, who was enrolled in Feb. 2009. He received his Bachelor of Engineering degree in Computer Science and Technology from Qingdao University of Science and Technology, P. R. China in 2005, and the Master of Engineering degree in Computer Software and Theory from Qingdao University, P. R. China in 2008. His current research focuses on the trust management and evaluation in online social networks. At Macquarie, Guanfeng’s supervisors are Dr. Yan Wang and Prof. Mehmet Orgun.
Award: Best Paper Award
Publication: Guanfeng Liu, Yan Wang, Mehmet A. Orgun and Ee-Peng Lim, A Heuristic Algorithm for Trust-Oriented Service Provider Selection in Complex Social Networks. 2010 IEEE International Conference on Services Computing (SCC 2010, ERA rank A) (research track, 29 papers accepted out of 165 submissions, acceptance rate=17.6%), pp. 130-137, Miami, Florida, USA, July 05-10, 2010
Seminar on 5 May, 2011
Speaker: Huiyuan Zheng
Title: QoS Analysis for Web Service Composition
The quality of service (QoS) is a major concern in the design and management of Web service compositions. Existing QoS calculation methods cannot deal with service compositions with complex structures and are time consuming when it comes to probability distribution modelled QoS. A systematic approach is proposed to estimate QoS for Web service compositions. Four types of basic composition patterns in a service composition are discussed. A set of formulae are developed to calculate the QoS probability distributions for the basic composition patterns. In particular, QoS solutions are provided for unstructured conditional and loop patterns. By using a directed graph to model a Web service composition, patterns are able to be identified recursively by a depth-first search (DFS) method so that complex structures in service compositions can be dealt with. Experimental results show that comparing with existing approaches, the proposed QoS estimation approach greatly improves the efficiency in probabilistic QoS estimation and can be used in large scale service compositions.
Biography: Ms. Huiyuan Zheng is a PhD candidate in Department of Computing, Macquarie University. She received a Bachelor of Engineering degree in Communication from Beihang University (China). Her research interests are in the areas of Service Oriented Computing and performance analysis of distributed systems. Her supervisor is Prof. Jian Yang.
Award: Best Student Paper Award
Publication: Huiyuan Zheng, Weiliang Zhao, Jian Yang and Athman Bouguettaya, QoS Analysis for Web Service Composition. IEEE International Conference on Services Computing (SCC 2009, ERA rank A), Bangalore, India, September 23-26, 2009, 235-242, (research track, 35 papers accepted out of 189 submissions, acceptance rate=18.5%)