Mining differential dependencies: A subspace clustering approach
Date
2014
Authors
Kwashie, S.
Liu, J.
Li, J.
Ye, F.
Editors
Wang, H.
Sharaf, M.A.
Sharaf, M.A.
Advisors
Journal Title
Journal ISSN
Volume Title
Type:
Conference paper
Citation
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2014 / Wang, H., Sharaf, M.A. (ed./s), vol.8506 LNCS, pp.50-61
Statement of Responsibility
Conference Name
25th Australasian Database Conference (ADC) (14 Jul 2014 - 16 Jul 2014 : AUSTRALIA, Univ Queensland, Brisbane)
Abstract
The discovery of differential dependencies (DDs) is the problem of finding a minimal cover set of DDs that hold in a given relation. This paper proposes a novel subspace-clustering-based approach to mine DDs that exist in a given relation. We study and reveal a link between δ-nClusters and differential functions (DFs). Based on this relationship, we adopt and co-opt techniques for mining δ-nClusters to find the set of candidate antecedent DFs of DDs efficiently, based on a user-specified distance threshold. Furthermore, we define an interestingness measure for DDs to aid the discovery of essential DDs and avoid the mining of an extremely large set. Finally, we demonstrate the scalability and efficiency of our solution through experiments on real-world benchmark datasets.
School/Discipline
Dissertation Note
Provenance
Description
Access Status
Rights
Copyright 2014 Springer International Publishing Switzerland