Generating top-k mediated schemas for multiple data sources

Guo Hui Ding*, Guo Ren Wang, Xiang Guo Zhao

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

Abstract

Schema integration is a critical step in many database applications, such as data space, data warehousing and electronic commerce, etc. This paper proposed an automatic approach to generate the mediated schemas over a set of source schemas. Firstly, the concept graph is presented to represent the source schemas for the unified representation. Secondly, the similarity between concepts is divided into intervals for the generation of the three merging strategies. Finally, the simulated annealing algorithm is employed to automatically generate the best k mediated schemas. Through extensive experiments, the results show that the algorithm proposed is effective and the running time is little.

Original languageEnglish
Pages (from-to)1700-1703
Number of pages4
JournalDongbei Daxue Xuebao/Journal of Northeastern University
Volume32
Issue number12
Publication statusPublished - Dec 2011
Externally publishedYes

Keywords

  • Concept graph
  • Mediated schema
  • Schema integration
  • Similarity
  • Source schema

Fingerprint

Dive into the research topics of 'Generating top-k mediated schemas for multiple data sources'. Together they form a unique fingerprint.

Cite this