Abstract

s . the part of the database to be mined (called the minable view ), . the type of pattern/rule to be mined, and . the properties that the patterns should satisfy. These patterns should include not only numerical constraints on statistical properties (like support, confidence, and correlation), but also those based on attribute domains, classes, and aggregates, such as "I.type = `snacks' and avg(I.price) < 100." Second, a data-mining system should support efficient processing and optimization of mining queries by providing a sophisticated mining-query optimizer. Such an optimizer exploits the various constraints stated in the user-specified mining query and their properties to generate access plans that guarantee a level of performance commensurate with the constraints in the query. CONSTRAINTS: ESSENTIALS FOR AD HOC DATA MINING We divide constraints into five categories: . Knowledge type constraints specify the type of knowledge to be mined, such as concept

Citations