Software fault feature clustering algorithm based on sequence pattern

Jiadong Ren*, Changzhen Hu, Kunsheng Wang, Dongmei Zhang

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

1 Citation (Scopus)

Abstract

Software fault feature analysis has been the important part of software security property analysis and modeling. In this paper, a software fault feature clustering algorithm based on sequence pattern (SFFCSP) is proposed. In SFFCSP, Fault feature matrix is defined to store the relation between the fault feature and the existing sequence pattern. The optimal number of clusters is determined through computing the improved silhouette of fault feature matrix row vector, which corresponds to the software fault feature. In the agglomerative hierarchical clustering phase, entropy is considered as the similarity metric. In order to improve the time complexity of the software fault feature analysis, the fault features of the software to be analyzed are matched to each centroid of clustering results. Experimental results show that SFFCSP has better clustering accuracy and lower time complexity compared with the SEQOPTICS.

Original languageEnglish
Title of host publicationWeb Information Systems and Mining - International Conference, WISM 2009, Proceedings
Pages439-447
Number of pages9
DOIs
Publication statusPublished - 2009
EventInternational Conference on Web Information Systems and Mining, WISM 2009 - Shanghai, China
Duration: 7 Nov 20098 Nov 2009

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume5854 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

ConferenceInternational Conference on Web Information Systems and Mining, WISM 2009
Country/TerritoryChina
CityShanghai
Period7/11/098/11/09

Keywords

  • Entropy
  • Improved silhouette
  • Sequence pattern
  • Software fault feature clustering

Fingerprint

Dive into the research topics of 'Software fault feature clustering algorithm based on sequence pattern'. Together they form a unique fingerprint.

Cite this