Research Seminars
Autumn 2009, Spring/Summer 2010
The department hosts a programme of research seminars by invited speakers and by the staff of the department. They are open to all. The following seminars will be taking place at the Department of Computer Science and Information Systems (DCSIS), Birkbeck College, or London Knowledge Lab (LKL), Birkbeck College and Institute of Education.
If you are interested in presenting a seminar please contact
- Susana Nascimento, Universidade Nova de Lisboa, Lisboa, Portugal.
- Subject: Additive Fuzzy Spectral Method for Cluster Analysis of Community Structure, Affinity and Relational Data, and Research Activities
- Date: Tuesday 9 March 2010, 17:00
- Location: Room 124 DCSIS Senate House
- Helen Karatza, Aristotle University of Thessaloniki, Greece
- Subject: Scheduling Policies for Distributed Systems
- Date: Tuesday 23 February 2010, 17:00
- Location: Room 124 DCSIS Senate House
- Karsten Maple, University of Bedford, UK
- Subject: Emerging Security Risks and Techniques for Authentication
- Date: Tuesday 9 February 2010, 17:00
- Location: Room 124 DCSIS Senate House
- Fatos Xhafa, Technical University of Catalonia, Spain
- Subject: Online Collaborative Learning in a Real Virtual Campus – The case of Open University of Catalonia (Barcelona, Spain)
- Date: Thursday 10 December 2009, 17:00
- Location: London Knowledge Lab, Large Seminar Room (base floor)
- Tassos Tompros, Queen Mary, University of London, UK
- Subject: Helping searchers with the query formulation problem: studies in Interactive Query Expansion
- Date: Tuesday 1 December 2009, 17:00
- Location: Room 124 DCSIS Senate House
- Dimitris Tasoulis, Imperial College London, UK
- Subject: Classification in Dynamic Streaming Environments
- Date: Tuesday 10 November 2009, 17:00
- Location: Room 124 DCSIS Senate House
- Fatos Xhafa, Technical University of Catalonia, Spain
- Subject: Middleware for P2P Collaborative Systems.
- Date: Tuesday 6 October 2009, 17:00
- Location: Room 124 SCSIS Senate House
Abstracts
Additive Fuzzy Spectral Method for Cluster Analysis of Community Structure, Affinity and Relational Data, and Research Activities
Susana Nascimento, Universidade Nova de Lisboa, Lisboa, Portugal.
Room 124 DCSIS Senate House - Tuesday 9 March 2010, 17:00
A spectral method for fuzzy clustering is proposed. The method operates on an extension of the spectral decomposition of a square matrix and extracts clusters one by one, which allows us to draw several stopping rules to the procedure. This applies to several relational data types differently normalized: network structure data (the first eigen-vector subtracted), affinity between multidimensional vectors (the pseudo-inverse Laplacian transformation), more conventional relational data, and in-house data of similarity between research topics according to working of a research center. This work was done in collaboration with Boris Mirkin.
Susana Nascimento received a Ph.D. degree in Computer Science from Universidade Nova de Lisboa in 2002. She is an Assistant Professor in the Department of Computer Science, Faculdade de Ci\^encias e Tecnologia, Universidade Nova de Lisboa, Lisboa, Portugal. Her main research interests include fuzzy cluster analysis and classification with applications to data mining problems in general and remote sensing imagery.
Scheduling Policies for Distributed Systems
Helen Karatza, Aristotle University of Thessaloniki, Greece
Room 124 DCSIS Senate House - Tuesday 23 February 2010, 17:00
Distributed systems are of great significance to meet the needs of demanding applications because they have increased computing power, are cost-effective and easily scalable. The most important aspect of a distributed system is the scheduling algorithm. It is responsible for the allocation of processors to jobs and determines the order in which jobs will be executed on processors. Distributed system scheduling has been studied extensively while applying different performance goals to many distributed system types and parallel workload types. Distributed computing environments such as heterogeneous clusters, multiple sites and grids, present new challenges for schedulers. The heterogeneity of these systems and the versatility of the resources often require special scheduling techniques for each group of resources in the distributed computing environment. This talk presents several scheduling policies employed in different distributed system types. Their performance is shown for various workload types.
Helen Karatza is a professor in the Department of Informatics at the Aristotle University of Thessaloniki, Greece. Her research interests include Computer Systems Modeling and Simulation, Performance Evaluation of Parallel and Distributed Systems, Resource Allocation and Scheduling, Cluster Computing, Grid Computing, Real-time Distributed Systems and P2P Systems. Dr. Karatza is a Senior member of IEEE, ACM and SCS. She is Editor-in-Chief of Simulation Modeling Practice and Theory (Elsevier), Area Editor of the Journal of Systems and Software (Elsevier) and Associate Editor of the ACM Transactions on Modeling and Computer Simulation.
Emerging Security Risks and Techniques for Authentication
Karsten Maple, University of Bedford, UK
Room 124 DCSIS Senate House - Tuesday 9 February 2010, 17:00
The proliferation of wireless networks, often accessible remotely via the Internet, couple with the increasingly diverse and sensitive data has led to significant challenges in assuring security. This seminar will consider emergent threats in information security and present new methods for information assurance, particularly in respect of authentication.
tbd
Online Collaborative Learning in a Real Virtual Campus – The case of Open University of Catalonia (Barcelona, Spain)
Fatos Xhafa, Technical University of Catalonia, Spain
London Knowledge Lab, Large Seminar Room (base floor) - Thursday 10 December 2009, 17:00
In this talk I will bring my experience as a teacher (since year 1999) of online distance studies at the Open University of Catalonia (Barcelona, Spain). The talk will focus on how we have approached the group collaborative learning in a real setting of the virtual campus, from both methodological and technological perspectives. The talk will discuss methodological aspects, technological aspects, and lessons learned, as well as research issues and challenges.
Fatos Xhafa was born in Albania and is Associate Professor at the Technical University of Catalonia and Open University of Catalonia, in Barcelona, Spain. One of his research interests is the use of distributed computing systems to assist the online distance learning paradigm. He is also interested in parallel and distributed algorithms, combinatorial optimization, and grid and P2P computing. Fatos has authored/co-authored over 100 papers in conferences, 50 peer-reviewed articles in international journals and 15 chapters in international books. He has also co-edited several books and conference proceedings. Please visit http://www.lsi.upc.edu/~fatos/ for more information on Fatos' academic activities.
Helping searchers with the query formulation problem: studies in Interactive Query Expansion
Tassos Tompros, Queen Mary, University of London, UK
Room 124 DCSIS Senate House - Tuesday 1 December 2009, 17:00
We use search systems in a variety of guises (e.g. Internet search engines, bibliographic databases), on a daily basis, for a wide range of information seeking tasks. From the searcher’s point of view, one of the most challenging steps in the search process is that of query formulation: how to select good search terms. This is especially so when we are searching for information in an area with which we are unfamiliar. In this talk I will discuss Interactive Query Expansion (IQE), a method by which systems can offer query formulation support by suggesting possible terms that can be added to the searcher’s query. I will specifically focus on some of our recent work on IQE, which looks into how we can design interface support that encourages searchers to engage more with IQE, especially when performing information seeking tasks on difficult, or unfamiliar, topics.
Dr. Tassos Tombros is a Lecturer at the Department of Computer Science, School of Electronic Engineering and Computer Science, Queen Mary University of London. He is working in the area of Information Retrieval (IR), with current focus on interactive IR understanding how people search for information, and how electronic systems can help them with the search process. Dr. Tombros has published in various areas of IR, including document clustering, automatic text summarisation, query expansion, structured document retrieval, and interactive IR. He holds a PhD degree in Information Retrieval from the University of Glasgow. For his doctoral thesis he received the 2003 BCS/CPHC Distinguished Dissertations Award. For more information on Dr Tombros’s academic activities, including a full list of publications, please see: http://www.dcs.qmul.ac.uk/~tassos
Classification in Dynamic Streaming Environments
Dimitris Tasoulis, Imperial College London, UK
Room 124 DCSIS Senate House - Tuesday 10 November 2009, 17:00
Modern technology has allowed real-time data collection in a variety of domains, ranging from environmental monitoring to healthcare. Consequently, there is a growing need for algorithms capable of performing inferential tasks in an on-line manner, continuously revising their estimates to reflect the current status of the underlying process.
In such circumstances the typical intrinsic assumption that the population being studied does not change over time is violated. Consequently the theoretical properties of any method established under this assumption cease to be valid. In this talk, we are concerned with the construction of a common framework for on-line and temporally adaptive classifiers, capable of handling the possibly drifting decision boundaries arising in streaming environments. We will demonstrate the scheme's effectiveness in both real and simulated streaming data.
Dimitris K. Tasoulis received his Ph.D. in Mathematics of Computers and Decision Making, from the Department of Mathematics, and the Department of Computer Science, at the University of Patras, Greece in 2006. He served as a Research Associate at the Institute for Mathematical Sciences, at Imperial College London, for 2 years until 2008. At 2008, he became a lecturer at the Mathematics Department of Imperial College. His research activities focus on various aspects of Machine Learning and Data Mining, with an emphasis on Data stream Analysis.
Middleware for P2P Collaborative Systems.
Fatos Xhafa, Technical University of Catalonia, Spain
Room 124 SCSIS Senate House - Tuesday 6 October 2009, 17:00
In this talk I will present some recent research work we have done at Technical University of Catalonia and Open University of Catalonia (In Barcelona, Spain) on Middleware for P2P Collaborative Systems. The talk will also address the potential use of such Middleware for the development of Groupware tools to support online group collaboration in Virtual Campuses.The tentative structure of the talk is the following: Introduction to some preliminary concepts about Middleware, P2P collaborative systems, overview of a few existing Middleware for P2P systems in the literature, introduction and main features of Jxta-Overlay, a middleware based on Jxta library of Sun MicroSystems, lessons learned and finally research issues and challenges.
Bio:Fatos Xhafa is Associate Professor (with tenure) at the Technical University of Catalonia and Open University of Catalonia (Barcelona, Spain). His main research interests include Parallel and Distributed Algorithms, Combinatorial Optimization, Grid And P2P Computing. He is also interested in research issues related to the use of distributed computing systems to the online distance learning paradigm. Fatos has authored/co-authored over 100 papers in conferences, 50 peer-reviewed articles in international journals and 15 chapters in international books. He has also co-edited several books and conference proceedings.
Fatos currently serves the editorial board of six peer-reviewed international journals, including International Journal of Autonomic Computing (IJAC), International Journal of Grid and Utility Computing (IJGUC) and International Journal of Computer-Supported Collaborative Learning (ijCSCL); he has also guest co-edited in several international journals. Fatos has served and is currently serving as PC Co-Chair / General Co-Chair of several international conferences and workshops. Please visit http://www.lsi.upc.edu/~fatos/ for more information on Fatos' academic activities.