| The Knowledge Collective (TKC) is a multi-layer, multi-agent framework for information management in an intelligent knowledge base that supports a collection of agents called MicroDroids. The MicroDroids provide information management capabilities through a variety of interfaces for experts, human users, and software components. This information is stored in a variety of internal structures (e.g., Java objects, rules, relational data). The main concept is that information is stored in a format that is natural to the type of information being maintained (e.g., data, metadata, ontologies, concept maps, lexicons, rules). The Knowledge Collective makes ontology-based information accessible to many end users, maintainable by domain experts and reusable by many users across many applications without their needing to know how or where the information is stored. The Knowledge Collective's first use is in version 4 of CIRCSIM-Tutor, an Intelligent Tutoring System developed by Martha Evens and her group at the Illinois Institute of Technology in Chicago, IL.; The uniqueness of The Knowledge Collective is in the combination of technologies for problem solving and the use of a multi-agent system where each MicroDroid manages its own information and works with other MicroDroids to solve specific problems. Each MicroDroid reasons about the information it is managing using an Ontology Inference Engine (ONTIE) that is capable of combining all of its internal information structures for the purpose of access, maintenance and problem solving, especially involving Qualitative Reasoning. |
