Within the scope of the project, we propose to
develop and test Knowledge Representation-related methodologies using
sub-domains of Medicine as test beds. Such methodologies are concerned
with the ordinary steps of ontology development – for example knowledge
acquisition, information organization, information visualization and,
evaluation – in addition to other relevant subjects, like information
security. The first stage of the project was planned in order to
approach the following topics:
- Knowledge organization:
a ontology covering the physiological aspects of human blood,
which will be gradually expanded to other blood sub-domains;
- Expert
knowledge acquisition: a corpus-based pragmatic study covering the
daily activities of experts in a blood bank institution;
- Knowledge
acquisition from texts: use of natural language processing techniques
in order to establish guidelines specifically designed for medical
blood-related literature;
- Visualization tools for evaluation:
development of tools for the visualization of information, which are to
be used in user-centered evaluations of ontologies;
- Information
security: study aiming to ensure the privacy and confidentiality of
biomedical information in large medical institutions from a user-centered
perspective.
The project activities are being conducted in
UFMG, HF and CoE, the latter during a sabbatical leave (CoE stands for
Center of Excellence in the Life Sciences and Bioinformatics, State
University of New York at Buffalo).