Abstract
The described data-knowledge coordination model is suitable for the integration of databases (DBs) and knowledge bases (KBs) under a distributed environment. Since data and knowledge are classified in different ways (data by their attributes and knowledge by their subjects), the model introduces the concept of context modules to realize dynamic coupling of the independently managed data and knowledge for various applications. It allows three kinds of hierarchies: data object hierarchies for data integration, knowledge object hierarchies for knowledge integration, and context module hierarchies for data-knowledge coordination. We discuss an object deputy-based method for realization of this model by Smalltalk. The interpretive nature of Smalltalk can realize dynamic links among multiple DBs and KBs. Through these links, the independently managed data and knowledge are imported into and treated as objects in Smalltalk, where data and knowledge are integrated by their respective deputy object hierarchies, which can be constructed by deputy mechanisms introduced by the authors. The context module hierarchies for coordination of the integrated data and knowledge are realized based on the model-view-controller architecture of Smalltalk. Realization of the data-knowledge coordination model in this way can achieve desirable features, including autonomy, multiple viewpoints and dynamic classification.