We propose a new functional framework for modeling, querying and reasoning about OLAP databases. The framework represents data (data cubes and dimensional hierarchies) and querying constructs as first-order and second-order functional symbols respectively. A polymorphic attribute-based type system is used to annotate the functional symbols with proper type information. Furthermore, semantic knowledge about the functional symbols, such as the properties of dimensional hierarchical structures and algebraic identities among query constructs, can be specified by equations which permits equational reasoning on equivalence of OLAP queries and generalized summarizability of aggregate views. Categories and Subject Descriptors H.2 [Database Management]: Languages General Terms Languages,Theory Keywords OLAP, query-language, summarizability
Ken Q. Pu