A New Framework of Very Large Knowledge-bases | from Database and Constraint Logic Programming Points of View

There are many knowledge representation and programming languages, and problem solvers for various applications. The increasing requirements of applications and the changes of computing environment require the cooperation and integration of existing languages and systems, because many applications cannot be necessarily represented by a single language, processed by a single problem solver and a single computing environment, and located in a single place. In this paper, we propose a new framework of heterogeneous, distributed problem solving, which corresponds to very large knowledge-bases consisting of multiple heterogeneous knowledge-bases and databases, for more complex applications. Our model consists of two concepts: agent and environment. An agent is an autonomous problem solver or a knowledge-base, which may be nested with an environment and multiple (heterogeneous) agents, where an environment provides a common space where agents communicate each other. We also show the representability and adaptability of this simple model by using some examples.