IT Technical Report Series for the Year 1999

[ 2009 ( 1 ) ][ 2005 ( 3 ) ][ 2004 ( 12 ) ][ 2003 ( 9 ) ][ 2002 ( 16 ) ][ 2001 ( 12 ) ][ 2000 ( 11 ) ][ 1999 ( 15 ) ]

Technical Reports

Intrinsic motivation in the computer science classroom

Technical Report No. : pb
By : P. Byrne

Psychology studies identify two main reasons why humans are motivated - extrinsic and intrinsic motivation. With extrinsic motivation, there is some reward (or punishment) which dominates the task; with intrinsic motivation, the task is undertaken for its own sake - the completion of the task being its own reward. The latter form is often found among computer professionals and followers who tend to get absorbed in tasks which fascinate or interest them in a manner beyond any external contingency. In a classroom setting, the presence of intrinsic motivation in students has a positive effect on both their learning experience and the extent of their learning - learning becomes a self-driven task, rather than a chore. Using research evidence which indicates the factors which are present in intrinsically motivation situations, this paper presents how such techniques might be included in computer science classes in order to enhance student learning.

Download the Full Paper

An Intelligent Online Tutoring Agent

Technical Report No. : cf
By : C. Fullard

Given the current popularity of the Internet, it is not surprising to see a number of educational institutions showing interest in providing courses which utilise an online environment. In order to achieve effective use of the Internet as a teaching tool, particularly with respect to the notion of Open Learning, certain features will be required from any online learning environment, e.g. flexibility, ease of use, etc. The Intelligent Online Tutoring Agent (IOTA) seeks to provide a learning environment that dynamically models the learners' abilities and streams course material to the learner using criteria based on comparing the learner-model to a concept-complexity/concept-significance model of the course material. Such a model will be generated automatically by the system, allowing the course providers to concentrate on the course content and leave presentation aside. A dynamic ranking system will be used to illustrate appropriateness of material. This concept-complexity/concept-significance model will also be applied to resources discovered on the WWW by an agent; this information can be used to supplement the regular course material.

Download the Full Paper

Expression Refinement Explained

Technical Report No. : sf
By : S. Flynn

This paper describes a calculus for the stepwise and piecewise refinement of expressions. It provides a means for the derivation of executable expressions from initial specifications. We take the view that a refinement calculus consists of: a specification language, which usually includes constructs which are non-executable, but is a "super-language" of a programming language; a refinement relation between specifications, which posesses particular properties necessary for the refinement of specifications in a stepwise and piecewise manner; and a set of laws determining how such refinements may proceed. A specification language of expressions is described, which includes construct for reasoning about undefined terms and non-deterministic expressions, as well as improper expressions. Examples are used to illustrate how and why each type of expression might arise. An outline is given of how refinement of specifications might proceed, and a refinement relation is defined using six actions.

Download the Full Paper

Error Patterns in Cooperative Program Development

Technical Report No. : jg
By : J. Griffith and C. Mulvihill

Despite the existence of aids for programming, errors in programming development remain, and often prove very costly. Much more work is required in error management. This paper shows how error classifications can provide useful error management information. A number of error classifications already exist. Along with looking at some of these existing classifications, two classifications of looping errors are presented in this paper. The motivation behind the developed classifications is to relate a task under development to an existing repository of programming information which is stored according to the developed classifications. The relation of such information may assist one or more users working on the same or similar programming tasks.

Download the Full Paper

Business Modelling to Support the Development of Integrated Process Support Systems in the Extended Enterprise

Technical Report No. : gl
By : G. Lyons and E. Dunican

Recent years have witnessed rapid advances in the areas of Business Process Reengineering(BPR) and Software Engineering. In order for BPR to work, successful deployment of Information Technology is imperative, with the lack of adequate IT support being cited as the single largest impediment in BPR projects. Furthermore, new business concepts such as extended enterprises and agile enterprises which build upon basic BPR call for even more complex software support in an ever-changing global business environment. This paper addresses these problems and proposes a framework which supports the design and development of information systems which support business processes in the extended enterprise.

Download the Full Paper

Achieving Alignment of Information Systems Implementation and Organisational Change

Technical Report No. : tg
By : T. Geraghty and G. Lyons

Download the Full Paper

Marketspace Strategy and the European Information Society

Technical Report No. : js
By : J. Sviokla and M. Quinn

The purpose of this paper is to articulate and speculate on the nature of the risks and opportunities that European companies will face after the deregulation of telecommunications in 1998. Whether or not executives realize it, the Internet and the World Wide Web have ushered in a globally standard, set of protocols and infrastructure for information transport, messaging, and document sharing. The impact of this critical mass of standards will be at least as profound as the coming of standard parts for manufacturing, radically altering cost structures and capabilities. Now is the time to start planning for these changes. We are in the midst of this change in the nature of competition and the shift from marketplace thinking to a marketspace thinking. Systematic research at the OECD has shown that high local phone tariffs and monopolistic or quasi-monopolistic control of access to the Internet keeps adoption low. The price elasticity of Internet access is high indeed. The higher the price, the lower the use.

Download the Full Paper

Managing Organisational Knowledge Assets

Technical Report No. : ky
By : K. Young

In this examination of the management of organisational knowledge assets, a distinction is made between Information Management and Knowledge Management, presented as the management of the organisational environment within which intellectual assets are fostered and encouraged. This supportive environment calls for computer systems that connect employees and facilitate the retrieval and transfer of stored knowledge. Knowledge is presented, not as a collection of facts, but as a dynamic, evolving concept, the creation and management of which involves many different processes. Three interdependent knowledge processes are identified: knowledge creation, knowledge use and reuse, and knowledge aggregation. As the support of each process requires different IT resources, a comprehensive IT infrastructure, containing several core capabilities, is required to achieve greater success in supporting the creation, use and aggregation of knowledge in pursuit of business value in different settings.

Download the Full Paper

A review of Fuzzy Logic and its Applications

Technical Report No. : gg
By : G. Gorman

This paper gives an overview of the theory and application of Fuzzy Logic. Classical set theory has been extended to deal with sets (information) which are by nature vague and/or not well defined. Interest is beginning to grow in this field because it lends itself more easily to formalisation from natural than traditional methods when vagueness is involved. More importantly, it works. It has been developed for many control applications in industry and it has also found its way into research applications such as pattern recognition, Keywords: fuzzy logic, vagueness, pattern recognition.

Download the Full Paper

The Application of Neural Networks to Environment Modelling: the Estimation of Sea Surface Temperature Differences

Technical Report No. : sr
By : S. Redfern and B. Ward

Bulk sea temperature, which is defined as the temperature of the top 1 metre of the sea, is a crucial variable used by climate modelling systems. Remote infra-red devices, however, can measure only to a depth of 100 micrometres, due to water's high absorption of this optical bandwidth. The remotely measurable skin region of the ocean has a temperature which is on average cooler than the underlying bulk waters by 0.5oC. The difference between the sea's bulk and skin temperature is defined as DT = TB - TS where TB and TS are the bulk and skin temperatures, respectively. DT is used to calculate TB from TS, which is measurable by satellites such as the ATSR (along-track scanning radiometer), AVHRR (advanced very high resolution radiometer), and MODIS (moderate-resolution imaging spectrometer), as well as by in-situ radiometric measurements. Since the direct measurement of DT over wide areas is prohibitively time consuming, several models that attempt to estimate it from sets of meteorological observations have been proposed. Due to the complexity of the factors governing the behaviour of DT, existing models fall short of predicting it to the required accuracy of 0.1oC for climate modelling. Artificial Neural Networks (ANNs) are ideally suited to modelling time series data which contains many input parameters, after a statistically representative number of training cases has been applied. This paper describes the application of in-situ radiometric and meteorological data to an ANN with the aim of predicting DT. Comparisons with other published models show that the ANN achieves the most accurate predictions, to within the desired error range.

Download the Full Paper

Design of a Virtual Store using Distributed Object Technology

Technical Report No. : dc
By : D. Chambers

This paper describes aspects of a current PhD research project in the area of distributed object technology. The main focus of the project is the use of this technology within the domains of Groupware and Multimedia applications. Groupware applications typically involve co-operation between several users and are often expressed in terms of operations on shared (and possibly distributed) objects e.g. shared document editors, virtual reality environments or indeed shared network based services. Multimedia applications involve the use and transport of special forms of program data e.g. images, sound and audio / video streams. The main hypothesis of this project is that distributed object technology can be used effectively within these application domains, but that further research is required to determine the best way to use and deploy the available technology standards. CORBA (Common Object Request Broker Architecture) has emerged as the de-facto industry standard for how remote objects can communicate and be deployed in a heterogeneous, distributed environment. Competing standards include Java RMI (Remote Method Invocation) and DCOM (Distributed Component Object Model) from Microsoft.

Download the Full Paper

Collaborative Filtering as a Means to Reduce Information Overload

Technical Report No. : sh
By : S. Hughes and C.O' Riordan

One fundamental problem that exists with the World Wide Web is the difficulty in volved in locating relevant material given the enormous quantities of information available. Existing systems try to tackle this problem by using spiders to locate, and indexes to provide access to, these resources. Although useful, there are a number of flaws associated with this approach. We describe an approach using collaborative (or social) filtering which attempts to overcome these flaws. Collaborative filtering automates the informal social system where one person recommends material to another, by recognising shared interests and recommending material based on shared interests.

Download the Full Paper

Information Retrieval and Filtering - An Overview

Technical Report No. : cor
By : C.O' Riordan and H. Sorensen

The areas of information retrieval(IR) and information filtering(IF) have become very active research domains. The problems created by the large increase of available online information, of which the vast majority is largely unstructured, have accentuated the need for effective mechanisms to separate the relevant information from the irrelevant. This paper reviews the main approaches and systems used in IR and in the newer field of IF. The paper also includes an overview of systems which utilise social or collaborative filtering techniques to deal with the problem of information overload.

Download the Full Paper

Review of Effective Image Processing Techniques of Mammograms

Technical Report No. : tod
By : T. O' Doherty

X-ray image processing is a computerised system which enhances the amount of detail visible on a digitised x-ray image. The effect of this technique in the diagnosis of breast cancer, where the detection of early malignant tumours is essential for effective treatment, is reviewed in this paper. Mammograms (breast x-rays) are initially enhanced by either increasing the contrast of suspicious areas or by removing background noise. Various mathematical methods are then applied to detect the individual tumours depending on whether the tumour appears as a microcalcification cluster or a mass. Several of these methods are investigated here. Preliminary results indicate that image processing can extract mammographic information which is not apparent by visual inspection. The processed image may be used to assist in mammographic interpretation with the potential to diagnose breast cancer at an earlier stage.

Download the Full Paper

The Application of Digital Techniques to the Detection and Extraction of Archaeological Earthwork Monuments from Aerial Photographs

Technical Report No. : sr1
By : S. Redfern and G. Lyons

Aerial photographs are a valuable resource for archaeological assessment, but their accurate interpretation is extremely labour intensive. The application of digital image processing techniques to the assessment process has, to-date, primarily focused on providing automatic measurement and correction of distortions inherent in oblique images. However, the monument discovery and mapping process has remained a task for the human operator. The automatic detection and mapping of archaeological evidence from aerial photographs poses significant challenges for digital image processing. The major sources of these difficulties are: the extremely low signal-to-noise ratio of most monuments visible in aerial photographs; the fact that earthwork monument boundaries do not adhere to any strict morphological constraints, other than the fact that many can be loosely defined as 'sub-circular' closed loops; the fact that nearly all monuments are visible only as thin boundary features, with no morphologically or texturally recognisable internal features; the presence of clutter - primarily modern day objects of relatively high contrast, such as walls, houses, trees, and roads; and monument damage and occlusion, both of which further add to the problem of incomplete boundary evidence. This paper presents techniques that we have developed for the automatic detection and boundary tracing of sub-circular archaeological monuments from aerial photographs. It quantitatively assesses the performance of these techniques, and compares this to the performance of existing image processing solutions when applied to the task under investigation. Our techniques are shown to represent a significant improvement within this problem domain over existing techniques.

Download the Full Paper
 

• Disclaimer
• Privacy
• Copyright
• Accessibility