IRC PhD Students
Chek Foung Tan
|Dr Yinshan Tang|
Gopi Shah Jadeja
Mohamed Al Farsi
Oghale Tessy Askia
Sarah Shah Ali
Selection of Completed PhD Projects
A Semiotic Approach to Enterprise Infrastructure Modelling: The Problem Articulation Method for Analysis and Applications
Simon B. K. Tan (completed 2006)
Abstract: In recent years, advances in information technology and enterprise engineering have revolutionised the way businesses are conducted. Enterprises are complicated organisations consisting of artefacts'(enterprise structure, activities, processes, information, resources, behaviour, goals, and constraints) that must be coordinated, planned and engineered to accomplish the enterprise goals. Enterprise engineering is an engineering approach comprised of techniques, methods, and infrastructures to aid the design, analysis, and implementation of an enterprise system. However, many modellers forgo the enterprise modelling methods and techniques and use localised ad-hoc and point solutions that are not amenable for enterprise infrastructure design. The current methods and techniques available are ill-equipped to model the holistic requirements of today's complex enterprises.
This research adopts the PAM (Problem Articulation Method), rooted in semiotics, a study of signs, symbols, language and information. PAM offers a suite of techniques, which enables the articulation and analysis of business, technical and organisational requirements, delivering a rigorous enterprise infrastructure to support the model. The method works by eliciting and formalising project planning and modelling (processes, activities, relationships, responsibilities, communications, resources, agents, dependencies and constraints), where the abstractions are mapped to represent the manifestation of the 'actual' enterprise. The proposed enterprise framework incorporates PAM and MEASUR (Methods for Eliciting Analysing and Specifying Users Requirements) a suite of semiotics methods to model the agent's ontological relationships, defining its meaning (semantics) to improve the analytical capability of the model and elicit the normative behaviours of agents to enrich the enterprise model. The PAM method has been further extended to examine the method relevance and significance in dealing with innovation in the modelling and planning of complex enterprise projects. The PAM conceptual model created is subsequently analysed, tested and verified using an off-the-shelf discreteevent simulation software, to be modelled dynamically. The simulation models are also validated by the domain experts from the respective organisations to assess the consistency, coherence and logical correctness of the model. This modelling approach contributed to the further development of the PAM theories, methods and techniques and represents a significant improvement to facilitate enterprise infrastructure planning and in so doing, exploit the technological, business and organisational effectiveness.
A Semiotic Approach for Modelling Semantics in Content Management
Jennifer Fu (completed 2007)
Abstract: A large amount of information from distributed sources has been produced by organisations for sharing and reuse in business networks. As a result, currently users of information often struggle to find the most relevant and appropriate information in an efficient manner. New requirements for personalised information in an interorganisational environment have arisen. This has motivated this PhD research project in the area of content management, which plays a fundamental role in supporting the configuration of personalised information.
Content management is a process of creating, structuring, organising and publishing content. Nowadays, it often deals with large amount of content in different forms and contexts. Therefore, configurations of appropriate information for individual users depend on efficient management of content with rich content semantics. Unfortunately, current practice has shown that the semantics of content is often overlooked and difficult to be modelled in the course of content management. The missing or inappropriate meaning of content causes significant difficulties in personalising information for individual users, because users miss a social link between their background knowledge and the new information to interpret. In response to this, several modelling techniques have been evaluated and ontology approves to be one of the most appropriate solutions. However, most of current ontological models cannot formally represent constraints, especially social constraints among concepts. Furthermore, there is lack of a content management method focusing on semantics modelling of content.
This PhD research addresses the issue of semantics modelling in the course of content creation, organisation and configuration. This work is underpinned by the theories of constructivism and semiotics from a subjectivist viewpoint. It takes social constraints of content into semantics modelling and aims to well represent content with precise semantics in a social context. A Semantics Modelling Method for Content Management (SMCM) is devised to enhance the modelling of content semantics in current practice. It models content from a human and social perspective in terms of ontology charts and extends the use of norms as social constraints. It also provides a lifecycle from the modelling of content semantics to technical design of content management systems. The method has been applied in a case study which was facilitated by a prototype tool for application and validation.
Empirical Assessment of Aspect- Oriented Programming and Coupling Measurement in Aspect-Oriented Systems
Marc Bartsch (completed 2007)
Abstract: Aspect-Oriented Programming (AOP) has gained an increasing amount of attention lately in the research community. It has been introduced with the claim that certain kinds of applications may be better maintainable. To date, there is still little empirical evidence that this claim is true. A similar lack of research can be observed for the related technology of coupling measurements. Coupling measures for aspect-oriented systems have been suggested in the literature, but their theoretical base as well as their usefulness remains questionable.
The thesis addresses both these issues. Coupling measurement was investigated by a detailed specification of a coupling framework allowing a rigorous approach to the specification, analysis and comparison of aspect-oriented coupling measures. Part of this assessment effort was a theoretical and empirical validation of 25 aspect-oriented coupling measures with the help of maintenance data collected from a medium-size program developed in industry. It could be shown that 15 coupling metrics can be considered indicators of maintenance effort. The data also shows that the size of an aspect is the best indicator.
Two experiments have been conducted which involved software professionals and students to investigate the effect of aspect-oriented programming on software maintenance. The results suggest that the use of aspect-oriented versions of certain design patterns have a beneficial impact on the number or errors that the subjects made. However, the results also indicate that in order to gain this benefit more effort needs to be invested into understanding the software systems. Both experiments serve as a methodological template for future experimental work about aspect-oriented programming.
Effective Project Planning Utilising Agent Technology and Semiotic Theory in a Virtual Organisational Domain
Holly D. Parsons-Hann (completed 2008)
Abstract: Although project success rates have slowly increased in recent years, less than 20% of projects are considered 'successful' - that is, within budget and time and meeting the stakeholders' requirements. One of the main reasons cited for project failure is the lack of quality requirements, such as the specified requirements not reflecting the actual needs of the stakeholders. This problem is exacerbated when combined with the new virtual organisational paradigm that consists of multiple stakeholders distributed over multiple geographic locations. Not only do stakeholder requirements have to be elicited, negotiated and prioritised, they also have to be documented in an understandable way with as little ambiguity as possible. Due to the different time zones, multiple languages employed and different business cultures and norms employed in virtual organisations, identifying the relevant stakeholders is made much harder, as is trying to reach a consensus on the requirements of a project. This thesis proposes a solution to this problem through the use of multi-agent technology based on an extended BDI (Belief, Desire, Intention) architecture. Each stakeholder is assigned an agent to enter their project requirements. Once completed, all agents negotiate with the other agents to achieve an agreed list of requirements that fulfil all stakeholders' needs. The thesis presents a detailed model of this solution, called Req-Neg, together with a prototype design and two case studies that were used to validate its effectiveness. What are academic challenges and contributions to the body of knowledge from this PhD research?