Applied knowledge processing is inspired to combine the advantages of artificial intelligence with software engineering, system architecture and database technology to acquire new fields of applications of computer science.
The main problem of modeling real world domains is their complexity. So, a paradigm must be found, that divides a whole domain into a set of small parts that are easier to handle both by the system and by the human experts. The models must allow an adequate representation but also an efficient implementation.
One purpose of our research at first is to develop a paradigm for representation based on a case-oriented methodology for knowledge acquisition that starts with the elicitation of concrete cases, abstracts them into prototypes or case-patterns, and further in decontextualized knowledge to be used in generic problem solving methods.
Based on this models the main focus consist in the uniform representation of knowledge with different abstraction levels and inference methods based upon them.
The applications at last use case-based reasoning in combination with episode-, concept- and generic methods.
- knowledge acquisition and representation with terminologies, ontologies, episodes, cases, schemes and generic knowledge
- knowledge models - integration of deductive, analog, inductive and case based inference methods
- architectures - uniform management and processing of knowledge bases of different abstraction levels
- architectural framework for integration of software components of different specialization levels
- modelling and representation of diverse application domains, implementation of efficient, knowledge based application systems