Ontologies are used to record knowledge; in large organisations these might include: equipment and products; personnel records; capabilities, business processes and plans; intelligence on rival organisations; historical records. Information can be extracted, using e.g. data mining, to enable automated business processes, forecasting and scenario planning, etc.
After construction, ontologies may need to be combined, evolved and extended to be kept up to date and debugged and repaired if faults are found. Currently, all of these tasks are done manually. This is a time consuming, error prone process requring good skills; further, there are difficulties in combining ontologies where they are created by different groups, as they are likely to have different vocabularies and underlying formalisms. It is therefore necessary to have tools for automated ontology management that integrates all the tasks.
Category Theory was developed from generalising transformations between algebraic structures in mathematics; on an intuitive level this has analogies with mappings between ontologies (which have some structure). Category Theory has been used in information flow theory and institution theory, both of which have been applied to ontology management.
This project aims to use Category Theory as a basis to provide automated assistance in the construction, combination and maintenance of heterogeneous collections of ontologies; and to use the Ontology Repair System towards repairing ontologies.