Liz is a postdoc at the University of Oxford. She is currently investigating ways to make computerised clinical guidelines such as PROforma more flexible. It is normally the case that the differing resources and constraints of different clinical institutions mean that guidelines cannot always be implemented in exactly the same way at different locations, and it may also be the case that a user (clinician or patient) may have preferences that affect a choice in a guideline; we are looking at ways in which we can integrate such flexible behaviour with the recommended guideline. A particular focus of this work is on the use of goals to help provide the necessary flexibility and so we are working on formalising a clinical goal ontology.
Liz is also continuing her PhD work on argumentation-based agent dialogues. This has lead to work (in collaboration with Anthony Hunter) on dialogues where only partial arguments (called enthymemes) are used. Enthymemes are of particular interest as they are ubiquitous in human interaction and so it is important to understand them in order to build agents that can converse naturally with humans. They are also useful for the protection of sensitive data, where there may be some information that the agent does not want to send across a network and so wishes to omit from an argument.
Another dialogue system that Liz has recently developed (in collaboration with Katie Atkinson) combines persuasion over actions with inquiry over beliefs. They have demonstrated this system with a medical example of a multi-agent system that comprises of a patient agent that represents a specific patient's data and a set of specialist medical agents that each have a particular value of concern.
For more details see Liz's personal page.
Email: lizblack at robots.ox.ac.uk