1999 · 133 citations · 4 references
ComputingComputer ScienceAlgorithmsData MiningDatabasesAggregation OperatorsData AggregationData WarehousingDimensionality ReductionDynamic ProgrammingOnline Analytical ProcessingDatabase Management SystemsGreedy HeuristicsInformation-theoretic ExplorationMulti-dimensional AggregationOlap AggregationOnline Analytical Processing (Olap) ArchitectureOperatorsSum Aggregation
OLAP systems currently provide only basic aggregates and navigation, forcing analysts to rely on intuition, which becomes tedious and error‑prone as data dimensionality and size increase. The authors aim to augment multidimensional databases with advanced mining primitives, introducing a new operator that supplies summarized reasons for aggregate increases or decreases. They develop an information‑theoretic formulation and a single‑pass dynamic programming algorithm that pre‑mines exceptions and summarizes reasons for aggregate changes, improving over a prior multi‑pass greedy approach. The approach eliminates manual drill‑downs, enabling analysts to see reasons for aggregate changes directly and achieves significant efficiency gains over the earlier greedy algorithm.
Our goal is to enhance multidimensional database systems with advanced mining primitives. Current Online Analytical Processing (OLAP) products provide a minimal set of basic aggregate operators like sum and average and a set of basic navigational operators like drill-downs and roll-ups. These operators have to be driven entirely by the analyst’s intuition. Such ad hoc exploration gets tedious and error-prone as data dimensionality and size increases. In earlier work we presented one such advanced primitive where we premined OLAP data for exceptions, summarized the exceptions at appropriate levels, and used them to lead the analyst to the interesting regions. In this paper we present a second enhancement: a single operator that lets the analyst get summarized reasons for drops or increases observed at an aggregated level. This eliminates the need to manually drill-down for such reasons. We develop an information theoretic formulation for expressing the reasons that is compact and easy to interpret. We design a dynamic programming algorithm that requires only one pass of the data improving significantly over our initial greedy algorithm that required multiple passes. In addition, the algorithm uses a small amount of
4
Discovery of Multiple-Level Association Rules from Large Databases
Jiawei Han, Yongjian Fu · 1995 · 923 citations
Statistical Analysis for Decision Making
Shimshon Kinory, Morris Hamburg · Technometrics · 1978 · 198 citations