Authors: Yao Wang; Zheng Qin; Fengfei Zhao; Jun Fang
Addresses: School of Software, Tsinghua University, Beijing city 100084, China ' School of Software, Tsinghua University, Beijing city 100084, China ' School of Software, Tsinghua University, Beijing city 100084, China ' School of Software, Tsinghua University, Beijing city 100084, China
Abstract: Usually, storage performance of data provenance is sustained through the use of compression algorithms. When provenance models with high repeatable rate and small variances are concerned, many current algorithms still face the challenge of avoiding redundancy. Adaptive Merge algorithm, featuring the optimisation of provenance factorisation, is introduced to compress provenance tree on the self-adaptive granularity basis. A kind of comprehensive solution regarding pointer-saving storage is provided here. The validity of Adaptive Merge algorithm is proven experimentally. Comparisons are made with Argument Factorisation to prove that our algorithm performs better in compression ratio. The extent of ratio enhancement depends on the features of data sets. The compression ratio of the provided provenance data set is increased by about 10%. Other data sets all show improvements in varying degrees. This suggests that by resolving pointer explosion issue, this new algorithm performs better and achieves a more reasonable compression ratio.
Keywords: data provenance; compression algorithms; provenance factorisation; self-adaptive granularity; computer applications.
International Journal of Computer Applications in Technology, 2013 Vol.47 No.4, pp.392 - 398
Published online: 02 Dec 2013 *Full-text access for editors Access for subscribers Purchase this article Comment on this article