Over the past few years, modern computing has shifted from problems of improving computer speed and disk capacity to methods of organising and managing the enormous volumes of information that we have access to.  It is not unusual for organisations to store petabytes of data each year. 

We have also moved from looking at the shape of data to the meaning of that data.  Whether it is downloaded from the Internet or an organisation's local data capture, there is little point in storing information if we can't make sense of it.

These two related problem areas are the basis for our research.  Data mining looks at mechanisms for finding nuggets of knowledge in a mass of data while conceptual modelling attempts to organise data in a manner that means that we retain its meaning and its context.  Our program of research thus has two interacting parts:

  • The Data Mining part focuses on investigates the development and application of advanced algorithmic, architectural and visualisation techniques to the problems of deriving, or assisting in the derivation of, knowledge from potentially large volumes of complex data.  These three areas use allied techniques and overlapping methodologies although they differ in the quantities of data, the initial information content of that data and the way in which the data is presented to the user.
  • The Conceptual Modelling part aims to develop conceptual models for complex systems and to explore how these models can be used to build data-oriented solutions to difficult problems.  In particular, while current conceptual modelling techniques have proved adequate for many problems, they cannot adequately model many advanced semantic relationships characteristic of complex systems.

These two areas come together when we investigate the accommodation of ontologies in data mining and in other ways.


Recent grant income

  • McIntyre, J., Roche, A., Roddick, J. F., Morgan, D. L. and Metcalfe, M. (2005-2007). Addressing Indigenous complex health, housing and social inclusion issues through critical systems approaches to build workforce capacity. SA Department of Human Services and Neporendi Aboriginal Forum Inc., ARC Linkage: $72,444 (ARC) + $24,000 (Partners) = $96,444.
  • Roddick, J. F., Ceglar, A. and de Vries, D. (2006-2007). Data Mining for Defence Information Analysis, DATS Phase 2. DSTO, Contractual Research: $104,000. Primary grant of $80,000 plus an additional $24,000 approved for an addenda to the project in 2007.
  • Roddick, J. F. and Calder, P. R. (2006-2008). Managing and Mining Evolving Ontologies through combining Patterns Languages and Data Mining. Royal Australian College of Surgeons and Australian Computer Society, ARC Linkage: $147,000 (ARC) + $49,500 (Partners).
  • Roddick, J. F. and de Vries, D. (2007). CrowData - Data Warehousing Project. Adelaide Football Club, Contractual Research: $5,000.
  • Roddick, J. F., Ceglar, A. and de Vries, D. (2007-2008). Detection of Anomalies in Transaction Sequences (DATS) Project and Other Related Research - DATS Stage 3. DSTO, Contractual Research: $120,000.
  • Bull, M. J., Roddick, J. F. and Sih, A. (2010-2012). Behavioural syndromes and social networks in sleepy lizards. ARC Discovery: $375,000.
  • Leinweber, D. B., Adelson, D. L., Bradshaw, C. J., Cooper, A., Denier, J. and Roddick, J. F. (2011). Enhancement of South Australian high-performance computing facilities. ARC LIEF: $430,000.
  • Roddick, J. F. (2011-2013). Techniques for Active Conceptual Modelling and Guided Data Mining for Rapid Knowledge Discovery. rL-Solutions, ARC Linkage: $163K (ARC) + $60,000 (Partners) = $223K.

Further information

We would be please to supply further information about our activities.  In particular, opportunities exist for high achieving postgraduates to join the program.  Please contact Professor John Roddick.