A Probabilistic Model for Data Cube Compression and Query Approximation

Download
  1. (PDF, 346 KB)
AuthorSearch for: ; Search for: ; Search for: ; Search for:
TypeArticle
ConferenceThe ACM Tenth International Workshop on Data Warehousing and OLAP, November 9, 2007., Lisbon, Portugal
AbstractDatabases and data warehouses contain an overwhelming volume of information that users must wade through in order to extract valuable and actionable knowledge to support the decision-making process. This contribution addresses the problem of automatically analyzing large multidimensional tables to get a concise representation of data, identify patterns and provide approximate answers to queries. Since data cubes are nothing but multi-way tables, we propose to analyze the potential of a probabilistic modeling technique, called non-negative multi-way array factorization, for approximating aggregate and multidimensional values. Using such a technique, we compute the set of components (clusters) that best fit the initial data set and whose superposition approximates the original data. The generated components can then be exploited for approximately answering OLAP queries such as roll-up, slice and dice operations. The proposed modeling technique will then be compared against the log-linear modeling technique which has already been used in the literature for compression and outlier detection in data cubes. Finally, three data sets will be used to discuss the potential benefits of non-negative multi-way array factorization.
Publication date
LanguageEnglish
AffiliationNRC Institute for Information Technology; National Research Council Canada
Peer reviewedNo
NRC number49870
NPARC number5763914
Export citationExport as RIS
Report a correctionReport a correction
Record identifieraede0cb1-b744-4692-a39b-af0315612a67
Record created2009-03-29
Record modified2016-05-09
Bookmark and share
  • Share this page with Facebook (Opens in a new window)
  • Share this page with Twitter (Opens in a new window)
  • Share this page with Google+ (Opens in a new window)
  • Share this page with Delicious (Opens in a new window)