A Multi-Agent method for parallel mining based on rough sets

Rough set is a relatively new AI technique in data mining. Multi-Agent system (MAS) has become a hotspot in the field of distributed AI recently. The challenge of the information age yet has not been resolved and the decision can't be made precisely and in time according to market and requirements. To improve the performing efficiency of data mining system, the paper defines the novel operations and reasoning of agents and a Multi-Agent method for parallel rule mining based on Rough sets is proposed. The information system is decomposed into many sub-information systems and every sub-information system can be an agent using rough set to acquire rules. From results of parallel mining, decisions can be made quickly and precisely.