Exploiting the multi-append-only-trend property of historical data in data warehouses

Hua Gang Li, Divyakant Agrawal, Amr El Abbadi, Mirek Riedewald

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

Data warehouses maintain historical information to enable the discovery of trends and developments over time. Hence data items usually contain time-related attributes like the time of a sales transaction or the order and shipping date of a product. Furthermore the values of these time-related attributes have a tendency to increase over time. We refer to this as the Multi-Append-Only-Trend (MAOT) property. In this paper we formalize the notion of MAOT and show how taking advantage of this property can improve query performance considerably. We focus on range aggregate queries which are essential for summarizing large data sets. Compared to MOLAP data cubes the amount of pre-computation and hence additional storage in the proposed technique is dramatically reduced.

Original languageEnglish
Pages (from-to)179-198
Number of pages20
JournalLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume2750
Publication statusPublished - 1 Dec 2003
Externally publishedYes

Fingerprint

Data warehouses
Data Warehouse
Historical Data
Freight transportation
Sales
Attribute
Query
Data Cube
Date
Large Data Sets
Transactions
Trends
Range of data

ASJC Scopus subject areas

  • Computer Science(all)
  • Biochemistry, Genetics and Molecular Biology(all)
  • Theoretical Computer Science

Cite this

Exploiting the multi-append-only-trend property of historical data in data warehouses. / Li, Hua Gang; Agrawal, Divyakant; El Abbadi, Amr; Riedewald, Mirek.

In: Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Vol. 2750, 01.12.2003, p. 179-198.

Research output: Contribution to journalArticle

@article{28ffcb4d95994d8bafe0d2979779f937,
title = "Exploiting the multi-append-only-trend property of historical data in data warehouses",
abstract = "Data warehouses maintain historical information to enable the discovery of trends and developments over time. Hence data items usually contain time-related attributes like the time of a sales transaction or the order and shipping date of a product. Furthermore the values of these time-related attributes have a tendency to increase over time. We refer to this as the Multi-Append-Only-Trend (MAOT) property. In this paper we formalize the notion of MAOT and show how taking advantage of this property can improve query performance considerably. We focus on range aggregate queries which are essential for summarizing large data sets. Compared to MOLAP data cubes the amount of pre-computation and hence additional storage in the proposed technique is dramatically reduced.",
author = "Li, {Hua Gang} and Divyakant Agrawal and {El Abbadi}, Amr and Mirek Riedewald",
year = "2003",
month = "12",
day = "1",
language = "English",
volume = "2750",
pages = "179--198",
journal = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
issn = "0302-9743",
publisher = "Springer Verlag",

}

TY - JOUR

T1 - Exploiting the multi-append-only-trend property of historical data in data warehouses

AU - Li, Hua Gang

AU - Agrawal, Divyakant

AU - El Abbadi, Amr

AU - Riedewald, Mirek

PY - 2003/12/1

Y1 - 2003/12/1

N2 - Data warehouses maintain historical information to enable the discovery of trends and developments over time. Hence data items usually contain time-related attributes like the time of a sales transaction or the order and shipping date of a product. Furthermore the values of these time-related attributes have a tendency to increase over time. We refer to this as the Multi-Append-Only-Trend (MAOT) property. In this paper we formalize the notion of MAOT and show how taking advantage of this property can improve query performance considerably. We focus on range aggregate queries which are essential for summarizing large data sets. Compared to MOLAP data cubes the amount of pre-computation and hence additional storage in the proposed technique is dramatically reduced.

AB - Data warehouses maintain historical information to enable the discovery of trends and developments over time. Hence data items usually contain time-related attributes like the time of a sales transaction or the order and shipping date of a product. Furthermore the values of these time-related attributes have a tendency to increase over time. We refer to this as the Multi-Append-Only-Trend (MAOT) property. In this paper we formalize the notion of MAOT and show how taking advantage of this property can improve query performance considerably. We focus on range aggregate queries which are essential for summarizing large data sets. Compared to MOLAP data cubes the amount of pre-computation and hence additional storage in the proposed technique is dramatically reduced.

UR - http://www.scopus.com/inward/record.url?scp=35248829149&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=35248829149&partnerID=8YFLogxK

M3 - Article

VL - 2750

SP - 179

EP - 198

JO - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

JF - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SN - 0302-9743

ER -