A Feature Ranking Algorithm in Pragmatic Quality Factor Model for Software Quality Assessment
Software quality is an important research area and has gain considerable attention from software engineering community in identification of priority quality attributes in software development process. This thesis describes original research in the field of software quality model by presenting a Feat...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English English |
| Published: |
2013
|
| Subjects: | |
| Online Access: | https://etd.uum.edu.my/3844/1/s804226.pdf https://etd.uum.edu.my/3844/7/s804226.pdf https://etd.uum.edu.my/3844/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Software quality is an important research area and has gain considerable attention from software engineering community in identification of priority quality attributes in software development process. This thesis describes original research in the field of software quality model by presenting a Feature Ranking Algorithm (FRA) for Pragmatic Quality Factor (PQF) model. The proposed algorithm is able to improve the weaknesses in PQF model in updating and learning the important attributes for software quality assessment. The existing assessment techniques lack of the capability to rank the quality attributes and data learning which can enhance the quality assessment process. The aim of the study is to identify and propose the application of Artificial Intelligence (AI) technique for improving quality assessment technique in PQF model. Therefore, FRA using FRT was constructed and the performance of the FRA was evaluated. The methodology used consists of theoretical study, design of formal framework on intelligent software quality, identification of Feature Ranking Technique (FRT), construction and evaluation of FRA algorithm. The assessment of quality attributes has been improved using FRA
algorithm enriched with a formula to calculate the priority of attributes and followed by learning adaptation through Java Library for Multi Label Learning (MULAN) application. The result shows that the performance of FRA correlates strongly to PQF model with 98% correlation compared to the Kolmogorov-Smirnov Correlation Based Filter (KSCBF) algorithm with 83% correlation. Statistical significance test was also performed with score of 0.052 compared to the KSCBF algorithm with score of 0.048. The result shows that the FRA was more significant than KSCBF algorithm. The main contribution of this research is on the implementation of FRT with proposed Most Priority of Features (MPF) calculation in FRA for attributes assessment. Overall, the findings and contributions can be regarded as a novel effort
in software quality for attributes selection. |
|---|