A kowledge-based model to database retrieval

Outside of knowledge management, databases represent the most knowledge intensive systems on our computers today. Yet, the knowledge embedded in these databases is tacit and remains so because, until now, there was no way to capture this knowledge and to integrate it into a suitable knowledge representation. We introduce a new knowledge structure and authoring system that bridges this gap. This knowledge structure organizes database content into a network of topic lists and paths that serves as a knowledge representation of how the database content is organized. The database schema provides the conceptual structure for this knowledge representation and the database content gives it meaning. Using automatic mapping tools, developers navigate over a database to capture lists of database values that serve as topics in the knowledge structure. They can also add topics and paths to the structure by hand. When this knowledge representation is complete, developers generate data files for a user navigation structure. In turn, this navigation structure enables users to browse and to explore database content and to pinpoint the information they need. Using this knowledge structure to organize content details paves the way for a new generation of information retrieval technologies that include agent-based capabilities.