Learn More
In this paper similarity relations and labeled block sets in incomplete information systems are introduced. Based on the two structures of granules, two rough set models are derived for mining of certain and possible rules in incomplete decision tables. The relationship between the two rough set models is examined.