Software defects are the key factors to evaluate the dependable software. This paper analyzes the attributes of software defects, and applies positive and negative association rules method to the research of software defects. This method can not only overcome the weak point of the traditional association rules method that can only mine the explicit rules, but also output some more meaningful rules of relationship of attributes. We extract the exterior relationship among the attributes of software defects, and fully mine the rules of inter- attributes. Through the application of the "Design and Implementation of Mining Linkage Management System of Coal Mine", the experimental results demonstrate that our mined rules have the advantages of less quantity, higher quality, fewer errors and conflicts.
[1]
Guan Xin.
Research of software defects classification
,
2008
.
[2]
Siddhartha R. Dalal,et al.
Using Defect Patterns to Uncover Opportunities for Improvement
,
1999
.
[3]
Jiawei Han,et al.
Data Mining: Concepts and Techniques
,
2000
.
[4]
Lu Yuchang,et al.
Study on Negative Association Rules
,
2004
.
[5]
Rajeev Motwani,et al.
Beyond market baskets: generalizing association rules to correlations
,
1997,
SIGMOD '97.
[6]
Hao Ke-gang.
Method of Software Defect Data Analysis and its Implementation
,
2008
.
[7]
Tim Menzies,et al.
Better Analysis of Defect Data at NASA
,
2003,
SEKE.