Affiliations: Dept. of Comp. Science, Univ. of North Carolina, Charlotte, N.C. 28223, USA. ras@uncc.edu | Informix Software Inc., 3350 West Bayshore Road, Palo Alto, CA 94301, USA. sucheta@informix.com
Abstract: A Distributed Knowledge-Based System (DKBS) is a collection of autonomous knowledge-based systems called agents which are capable of interacting with each other. A query can be submitted to one agent or a group of agents. An agent when contacted by the user acts as a master agent. If he is unable to answer the query, he looks for help from other agents which act as his slaves. In this paper, an agent is represented by an information system (either complete or incomplete), a collection of rules called a knowledge base and the Query Approximate Answering System (QAAS). Rules are interpreted as descriptions of some attribute values in terms of other attribute values. These descriptions are usually not precise and they only provide a lower approximation of attribute values. We say that an attribute value is reachable by an agent if either it belongs to the domain of one of the attributes in his information system or it is a decision part of one of the rules in his knowledge base. In the second case, we assume that all attribute values from the classification part of a rule have to be reachable. When rules are discovered by one site of DKBS which currently acts as a slave, they are sent to the master agent of that slave. The QAAS of the master agent will use these rules to answer a query submitted by the user and next it will store these rules in the agent's knowledge base. So, the set of reachable attribute values at any site of DKBS is constantly changing. Knowledge bases built that way might easily become inconsistent because rules they contain are created independently at different sites of DKBS. The problem of repairing inconsistent rules was investigated in [19]. In this paper, we propose a strategy for discovering rules in incomplete information systems and give a formal system for handling queries in DKBS where each site contains either an incomplete or a complete information system.
