The Artificial Intelligence Group

The objective of the Artificial Intelligence Group is to apply artificial intelligence techniques, in particular tempora llogic and case-based reasoning, to common-sense knowledge representation and reasoning, knowledge management and reuse, as well as real-life industrial, commercial and legal problems, such as pattern recognition, planning and metal casting.

Over the last 2 decades, the group has established a high-impact body of research work by focusing on temporal representation and case-based reasoning, including:

  • A general temporal theory, which treats both points and intervals as primitive on the same footing, thus can subsume other time theories/models.
  • A temporal database system, which allows both relative and absolute temporal knowledge and supports both valid time and transaction time.
  • A theory of action/event, which, in terms of a real reified temporal logic, formally specifies the so-called most general temporal constraint that guarantees the common-sense assertion that "beginning of the effect cannot precede the beginning of the cause", and characterises the most versatile causal relationships between actions/events and their effects.
  • An ontology of time, which addresses fundamental considerations of time, meta-predicates and temporal propositions. It can subsume those representative categories in the literature and provides the expressive power for modelling typical temporal terms/phenomena (e.g.: "starting-instant", "stopping-instant", "dividing-instant", "instigation", "termination" and "intermingling" etc.).
  • A historical case-based reasoning framework, which provides formal schemas and techniques for representing and recognizing scenario/history patterns.

The case-based reasoning research carried out at Greenwich is among the most prolific ones in this area in the UK and specialises in the application of case-based reasoning in complex problems involving structural similarity metrics. The group is also active in research in business intelligence and optimisation, as well as applications of artificial neural networks and genetic algorithms involving real-life industrial applications.

In particular, the following 2 examples provide evidence of research carried out by the AI group that not only benefits the theoretical research in the domain of Artificial Intelligence, but also has a high impact on society in the application areas of medical diagnosis and treatment, government spending and law enforcement:

  • The collaborative research project with Nepal Heart Foundation (NHF) was covered by Nepal's national newspaper, The Himalayan Times (23 August, 2015). According to NHF: "the system is the first of its kind in Nepal and will not only offer assistance to rural health professionals but also allow diagnosis of ARF at an early stage, and will cut the patients' costs and reduce government's financial burden".
  • The visualised case tool, "Visual Time", is an implementable system that is able to solve real-world problems. The technology would be of particular interest to the police and security services in investigating complex crimes. The tool would provide considerable cost savings to investigations by enabling large amounts of evidence and intelligence to be analysed. The application areas may be extended to other firms/organisations in the domain of planning/prediction, diagnosis/forecast, medical record, pattern recognition, etc.

The Artificial Intelligence Group at Greenwich currently consists of 6 research-active members.

For further information on the Artificial Intelligence Group, please contact Dr Jixin Ma (E-mail: j.ma@greenwich.ac.uk).

Centre for Computer and Computational Science is part of the Faculty of Architecture, Computing & Humanities, University of Greenwich.