Logo image
Back
Chapter 68 - A One-Pass Aggregation Algorithm with the Optimal Buffer Size in Multidimensional OLAP
Book chapter

Chapter 68 - A One-Pass Aggregation Algorithm with the Optimal Buffer Size in Multidimensional OLAP

Young-Koo Lee, Kyu-Young Whang, Yang-Sae Moon and Il-Yeol Song
Proceedings 2002 VLDB Conference, pp 790-801
2002
DOI: https://doi.org/10.1016/B978-155860869-6/50075-5

Abstract

This chapter considers an aggregation method that uses dynamic multidimensional files adapting to skewed distributions. Aggregation is an operation that plays a key role in multidimensional OLAP (MOLAP). Online analytical processing (OLAP) is a database application that allows users to easily analyze large volumes of data in order to extract the information necessary for decision making. OLAP queries make heavy use of aggregation for summarizing data, because summarized trends derived from the records are more useful for decision making rather than individual records themselves. Because computing aggregation is very expensive, good aggregation algorithms are crucial for achieving performance in OLAP systems. OLAP is based on a multidimensional data model that employs multidimensional arrays for modeling data. The multidimensional data model consists of measures and dimensions: measures are the attributes that are analyzed; dimensions are the attributes that determine the values of the measures. A dimension is mapped to an axis of the multidimensional array. A measure is mapped to a value stored in a cell. This model allows OLAP users to analyze changes in the values of the measures according to changes in the values of the dimensions.

Metrics

14 Record Views

Details

Logo image