DEPARTMENT OF COMPUTING SCIENCE AND MATHEMATICS
UNIVERSITY . COMPUTING SCIENCE . SEMINARS

SEMINARS - Autumn 2004

[Talk Schedule] [Abstracts] [Previous Seminars]

The Department of Computing Science and Mathematics presents the following seminars. Unless otherwise stated, seminars will take place in Room 4B94 of the Cottrell Building, University of Stirling from 15.00 to 16.00 on Friday afternoons during semester time. For instructions on how to get to the University, please look at the following routes.

If you would like to give a seminar to the department in future or if you need more information, please contact the seminar organisers, either Julie Cowie (Phone 01786 467446, Email jco@cs.stir.ac.uk) or David Cairns (Phone 01786 467445, Email dec@cs.stir.ac.uk).

Talk Schedule [Top] [Abstracts]

Abstracts [Top] [Schedule]

24th September[Schedule]

An Evaluation of RSVP Control Message Delivery Mechanisms
Professor Joe Sventek
Department of Computing Science, University of Glasgow

Abstract:

Although originally proposed for resource reservation in Internet Protocol integrated service networks, the Resource Reservation Protocol (RSVP) has been attracting considerable attention as a signalling protocol for use in multi-protocol label switching (MPLS) networks. Recently, MPLS signalling protocols, along with IP routing protocols, are being standardised as a control plane for packet, time-division, wavelength, and spatial switching networks. Generalised MPLS (GMPLS), as this control plane is known, will therefore potentially use a suitably enhanced form of RSVP. Of critical importance for the control plane for such networks is the resilience of the signalling protocol to loss of control packets. This talk compares a number of RSVP control message delivery mechanisms in terms of several performance metrics. A key discovery is that the loss of Path messages is typically more detrimental than the loss of Resv messages. As a result, we can confidently recommend modifications to the RSVP protocol that improve the resilience to signalling errors in GMPLS while optimising other performance parameters of the protocol.

1st October[Schedule]

67 Software Projects 1962-2002; what can we learn?
Professor Tim Denvir (Honorary Professor)
Department of Computing Science and Mathematics, University of Stirling

Abstract:

The term “Software Engineering” was coined in 1968, but the activity itself had been going on for a considerable time before that. Some differences are easy to recognise, but other trends are harder to identify. In the early 1960s the driving cost of software development was computer time, with the programmer’s labour a small, almost ignorable, fraction of the whole. It has now been the other way round for a long time. The tasks we program computers to carry out have changed a great deal. The tools, techniques, theoretical underpinnings, and issues of concern have all developed. The kinds of knowledge and skills needed by a software engineer are probably changing. I plan to give a quick survey of the sort of projects I have experienced over the years and see whether we can perceive any evolutionary patterns.

8th October[Schedule]

Families of Web Pages
Mr Richard Bland
Department of Computing Science and Mathematics, University of Stirling

Abstract: 

Most sets of WWW pages have some element of family structure. For example the Computing Science home page can be conceptualised as a child of the University home page, and the parent of the 31Q5 home page; the 31N5, 31P5 and 31Q5 home pages could be thought of as siblings. This seminar will present a method for generating such sets of web pages. The method ensures that each page has correct links to other members of the family. Adding new pages to the family, or reconfiguring the family, is a matter of editing one file and triggering a rebuild. The method can be used to construct both static pages and dynamic pages: as examples the seminar will discuss how HTML (static) and JSP (dynamic) pages can be generated. The seminar will also show how the method can be integrated with other techniques for managing house style: in particular, the use of XSLT to separate page content from page formatting.

15th October[Schedule]

Model Driven Architecture: A Pattern-Oriented Approach
Mr Alasdair McLaren
Pivotal Integration

Abstract: 

The Object Management Group is currently promoting its flagship project, Model Driven Architecture (MDA); a multi-platform modelling specification which addresses the complete life cycle of software projects from design through deployment, integration and management. MDA enables organizations to integrate new and existing solutions while allowing for expansion with future developments. It also promotes platform independence across all phases of the Software Development Life Cycle. A Software Design Pattern is an implementation of software best practices providing a well proven, generic solution to a classification of common software problems. In this seminar I intend to investigate MDA standards and principles in the context of best practice software solutions to determine the feasibility of a Pattern Oriented approach to software design. I also ascertain the benefits of such an approach.

22nd October [Schedule]

How do we database all the other bioscience data?
Dr Fred Howell
Institute for Adaptive and Neural Computation, University of Edinburgh

Abstract:  

Most of the expensively gathered experimental data in bioscience ends up on CDs gathering dust in cardboard boxes in labs, with cryptic filenames like "GX_0967rat179.dat" which mean little to anyone apart from the researcher who did the experiments.

What can be done about this? In this talk I'll discuss how one might move people on from the tools they actually use (long filenames, Word, Excel) without making them learn tools they'll never be able to use (relational databases) in the hope of getting scientific data out onto the "syntactic web" where it will be useful to others.

5th November [Schedule]

Ubiquitous services: harmony or discord?
Professor Evan Magill
Department of Computing Science and Mathematics, University of Stirling

Abstract:
Software services and applications controlling a wide array of devices and appliances offer exciting opportunities. But will such an array of controlling software and networked devices operate in a coherent and expected manner? Experience in large communications systems suggest not. Using the automated home as an exemplar, this talk describes a novel approach to avoid such chaos.

12th November[Schedule]

Extended Static Checking
Mr Thomas Wilson
Department of Computing Science and Mathematics, University of Stirling

Abstract:
In this seminar I will talk about Extended Static Checking (ESC) which is a technique that can be used to detect a range of possible run-time errors (array bounds errors, null dereference errors etc) at compile-time. ESC lies between conventional type checking and full formal specification and verification, allowing more problems to be detected than type checkers can but not requiring the full investments that are essential for the use of heavyweight formal methods. ESC tools feel like simple type checkers to use but employ advanced formal verification techniques behind-the-scenes. After introducing ESC, I’ll give a live demo of an ESC tool for Java called ESC/Java2, using it to check some simple examples. I’ll conclude by highlighting some problems that arise when using ESC/Java2 within large projects and explaining how my own system, Omnibus, helps tackle these by using a cleaner language as its basis and supporting the integrated use of ESC, full formal verification and Design by Contract within the same project. Could this kind of work form the basis of future software development? With most commercial companies that come into contact with ESC/Java2 showing great enthusiasm towards it and Microsoft recently joining the party with the development of their new Spec# system, I believe that it just might.

19th November [Schedule]

Innovative Use Of Technology To Support People With Dementia And The Dilemmas Of Using Monitoring Technology To Support People With Dementia
Colm Cunningham and Professor Mary Marshall
The Dementia Services Development Trust, University of Stirling

Abstract:
The Dementia Services Development Centre (DSDC) is based at the Iris Murdoch Building, part of the Department of Applied Social Sciences within the University of Stirling. The DSDC is designated as a centre of excellence on dementia care by the Scottish Executive and has a key role in providing training, information and consultancy. The centre has therefore developed considerable experience in the use of technology to support people with demenita and it hosts the Scottish Dementia Technology Group. Professor Mary Marshall will begin this seminar by drawing on the centre's experience of technology used in practice and will describe the potential of technology to support people who have dementia within the community and care settings. Professor Marshall will highlight the types of technologies available and the experience of using these in practice for people with dementia. Colm Cunningham will then consider the ethical and legal implications related to using technology. This will be presented in the context of a case study. The case study relates to a woman with dementia where technology was considered because her walking activity challenged their care home setting. He will describe some of the procedural, ethical and legal issues raised by this particular case.

26th November [Schedule]

Spike Detection
Dr. Nhamo Mtetwa
Department of Computing Science and Mathematics, University of Stirling

Abstract:
Neurophysiologists often record the currents flowing across cell membranes using insulated wire probes whose exposed tip is placed in the extracellular medium near the soma or other spiking region of one or more neurons. Any such probe will record currents passing across all sorts of physiological membranes close to its tip, but we can approximate this as a small number of close, high-amplitude sources (which we call foreground sources) imposed on a background of synaptic, somatic and axonal currents from more distant sources. Spike detection is the process of identifying the waveforms associated with action potentials in the foreground cell(s). The ways in which the foreground signals are mixed, and the size of the background signal result in the recorded signal being difficult to interpret. In this talk we discuss different thresholding criteria and different smoothing techniques to achieve spike detection.

Previous Seminar Series [Top] [Abstracts] [Schedule]

Last Modified: 24th September 2004