A Model for Detecting Cost-Prone Classes Based on Mahalanobis-Taguchi Method

Hirohisa AMAN  Naomi MOCHIDUKI  Hiroyuki YAMADA  

IEICE TRANSACTIONS on Information and Systems   Vol.E89-D   No.4   pp.1347-1358
Publication Date: 2006/04/01
Online ISSN: 1745-1361
DOI: 10.1093/ietisy/e89-d.4.1347
Print ISSN: 0916-8532
Type of Manuscript: Special Section PAPER (Special Section on Knowledge-Based Software Engineering)
metrics,  cost-proneness,  prediction,  discriminant analysis,  Mahalanobis-Taguchi method,  

Full Text: PDF>>
Buy this Article

In software development, comprehensive software reviews and testings are important activities to preserve high quality and to control maintenance cost. However it would be actually difficult to perform comprehensive software reviews and testings because of a lot of components, a lack of manpower and other realistic restrictions. To improve performances of reviews and testings in object-oriented software, this paper proposes a novel model for detecting cost-prone classes; the model is based on Mahalanobis-Taguchi method--an extended statistical discriminant method merging with a pattern recognition approach. Experimental results using a lot of Java software are provided to statistically demonstrate that the proposed model has a high ability for detecting cost-prone classes.