Information Retrieval from an Incomplete Data

A complete data cube is a data cube in which every aggregate value in the multidimensional space is stored or can be computed. An incomplete data cube is a data cube in which points in the multidimensional space are missing and cannot be computed. This paper describes an incomplete data cube design. An incomplete data cube is modeled as a federation of cubettes. A cubette is a complete sub-cube within the incomplete data cube. The incomplete cube is built piecemeal by giving a concise, high-level speciication of each cu-bette. An eecient algorithm to retrieve an aggregate value from the incomplete data cube is described. When a value cannot be retrieved because it is missing, alternatives at a lower precision that can be retrieved are iden-tiied. When a value can be partially computed (i.e., some of the values lower in the hierarchy are missing, but some are present) a measure of the completeness of the result is supplied along with the partially aggregated value. The design also includes an algorithm that removes redundant cubettes and an algorithm to increase the retrieval power of the federation through the creation of virtual cu-bettes. Permission to copy without fee all or part of this material is granted provided that the copies are not made or distributed for direct commercial advantage, the VLDB copyright notice and the title of the publication and its date appear, and notice is given that copying is by permission of the Very Large Data Base Endowment. To copy otherwise, or to republish, requires a fee and/or special permission from the Endowment.