• Macquarie Home »
• Faculty of Science »
• Department of Computing

Debbie Richards's Projects

Pedagogical planning in an Intelligent Agent-Augmented Multi-User Virtual Environment

The use of games and game engines to create learning environments is considered by many educationalists to be an effective way of increasing student motivation and learning. Game elements such as goals and competition can potentially increase student motivation, however, ensuring that the learning objectives are achieved is more challenging than being the first to finish a game.

This project will extend the prototype proof-of-concept Virtual Singapura (VS) Intelligent Agent-Augmented Multi-User Virtual Environments (IAA-MUVE), developed at Nanyang Technological University, so that the learning experience can be monitored and facilitated by the system to ensure successful learning outcomes. This will be achieved by using techniques from artificial intelligence and intelligent/adaptive user interfaces such as agents, knowledge-based systems, ontologies and user modelling. Some of the intelligence to be incorporated will be mined from existing data gatherine in a pilot study conducted in 2007 in Singapore. A technique which incrementally acquires and updates a mdoel of the knowledge of the learner and the learner experience will be developed.

Investigating the role of realism and empathy on knowledge transfer

Just as it is common in education to use a teacher, instructor, tutor, peer or group to guide and motivate the learner, eLearning environments may include a pedagogical agent to perform one or more of these roles. These agents have social ability based on the observation that learner's perception of the learning experience is positively affected when a lifelike character is included in a computer-based interactive learning environment. This has become known as the persona effect (Lester et al. 1997). Human qualities such as empathy (McQuiggan and Lester, 2006), enthusiasm and interesting personalities (Elliot, Rickel and Lester, 1999) and expressiveness in terms of communication and levels of advice (Lester et al 1997) have been perceived by learners and educators to be useful. A study by Moreno et al. (2000) found that a participant's memory retention and knowledge transfer was better when the learner's was assisted by a pedagogic agent compared to a computer-based text environment without the agent.

Increasing the realism of virtual agents is the goal of many research projects. However, it is well-known that as humans we willingly and frequently suspend disbelief and accept artificial characters and robots as if they embodied "life", even becoming emotionally involved with them and affected by them. Empathic agents go a step further, they appear to understand how we, the users, are feeling. Use of such agents in education is a new and challenging direction. This project will explore the role that realism plays with a particular focus on empathic agents on learning transfer.

Crime Scene Investigation

Assist police in honing and testing their skills in crime scene investigation. Noticing relevant objects and patterns in a crime scene is crucial for solving many crimes. Incorporate virtual reality technology with pattern matching and reasoning systems from artificial intelligence to assist police to train its officers. This project will involve working with the NSW Police Department. Desired prerequisite knowledge: COMP330 and COMP329.

An intelligent/adaptive user interface for training

Authoring of trainee systems is typically limited to trained specialists such as programmers. Similarly the acquisition of knowledge in a domain often requires the mediation of a trained knowledge engineer. This means that the domain expert and trainer needs to pass on what they know to another. This is time consuming, costly, error-prone and makes even small changes to the system problematic. This project aims to allow the trainer to change the knowledge and the training scenarios while interacting with the system. This project will involve extension of a knowledge acquisition technique known as ripple down rules and is part of a larger project being conducted with members of the Intelligent Systems Group, Visor and Language Technology Group in the computing Department. A prototype system created in Unreal Tournament will provide an initial platform to work on. Desired prerequisite knowledge: COMP330 and COMP329.

Working in team-based ICT projects

This project will build on work that began with Brook's famous paper "the Mythical-man month". When that paper was written, system development concepts and techniques such as object orientation and XP were not known or in widespread use. Today globally distributed teams are common in many ICT projects and the size and complexity of software continues to increase. Thus the use of teams seems even more necessary. Despite this there are many problems with working in teams. This project will draw on literature from multiple disciplines but focus on ICT teams and their peculiar needs and features. It is envisaged that a number of case studies will be conducted, both with student teams and in industry, to further explore the issues that arise and suggest some methods for improvement. One key issue to be considered is equitable distribution and reward of effort. Desired prerequisite knowledge: ISYS346/COMP345/COMP340, ISYS302.

Automatic validation of UML models

A key problem that many analyst/programmers, particularly students, have with UML is validation and cross-checking of the models. Like compiled code that can be executed with various inputs and checked against expected outputs, this project could explore how test cases and suites could be created to allow execution and automatic validation of UML. Alternatively, the project could involve evaluation of a range of UML modeling tools to suggest which provide the greatest support and further improvements. In any case, an expected outcome of the project would be a comprehensive model which allows the links between the different views created by UML, that is functional, structural and behavioural models, to be clearly understood and applied.

Do we need girls in IT?

The gender imbalance in the ICT workplace is well-recognised. Studies have been conducted which report various factors, such as the females cultural background and the perception of IT as a male field, influencing the decision to pursue a career in IT. To explain the gender imbalance various theories have been put forward. One theory is essentialism which holds that females are fundamentally different from males at the biological, pysiological and/or psychological level which makes them less suitable for and/ or attracted to technology (which is sometimes taken to mean mathematics and science in general).

A more popular theory is known as social construction in which IT is seen as a male domain. As neither theory fully explains the data collected in numerous gender and IT studies, in recent times Individual Differences Theory has begun to emerge which takes into "account the uniformity of social shaping messages conveyed in a culture but also takes into account the varied influence of individual background and critical life events that result in a range of responses tothose messages" (Trauth 2006, p.1156). This project will conduct studies to further understand the factors which affect career decisions in IT and contribute to developing theory in the area of gender and IT.

Reconciling Use Case Descriptions from Multiple Viewpoints

The design of a system will always be limited by a poor requirements specification. If the requirements are inconsistent, incomplete or invalid then the design will be inappropriate or even useless. Use cases or user stories are becoming increasingly popular as a means of capturing system requirements from the user's point of view. This process involves the development team brainstorming the main chunks of functionality and then describing the steps involved in each use case. Even though this is a group activity it is common for a few to dominate the group which results in incomplete use case descriptions that do not represent the requirements of the whole group. In the approach, known as RECOCASE because we use a CASE tool to assist requirements RECOnciliation, the group identify and then asynchronously capture multiple requirements viewpoints. The use case descriptions are entered in natural language and automatically processed by our system to produce a concept lattice. The concept lattice is used to reveal common ground and differences between the stakeholders. The group process enables conflicts to be identified and resolved, where possible, using our negotiation strategies and resolution operators. This approach will be further investigated and evaluated within real organisation as part of this project.