Users Review’s on Software Defect Prediction Utilizing Machine Learning methods

Document Type : Original Article

Authors

1 Department of Information Systems, Faculty of Computers and Artificial Intelligence, Helwan University, P.O.Box 11795, Egypt.

2 Department of Machine Learning and Information Retrieval, Faculty of Artificial Intelligence, Kafrelsheikh University, Kafrelsheikh, Egypt.

3 Department of Electrical Engineering, Faculty of Engineering, Kafrelsheikh University, Kafrelsheikh, Egypt

Abstract

Software Defect Prediction (SDP) is a crucial and helpful method for upgrading software reliability and quality. It enables more effective project management by predicting potential release delays early on and facilitating cost-effective corrective actions to enhance software quality. This is achieved by forecasting which modules in a large software product are likely to have the highest number of defects in the next version. However, creating reliable defect forecasting models remains a challenging issue, leading to the presentation of numerous methods in literature. Typically, machine learning (ML) classifiers are employed, using manually designed attributes (like complexity measures) to identify problematic code. However, these attributes often fail to capture the full structural and semantic details of the software. Incorporating this information is crucial for the development of accurate defect prediction models. This study covers various defect prediction strategies and explores recent research on ML methodologies for SDP, aiming to bridge the gap between software semantics and defect forecasting attributes. By doing so, it seeks to produce more precise and accurate forecasting.
 

Highlights

 

 

Keywords

Main Subjects

Full Text

 

 

 

 

References

 
 

Files

Share

How to cite

Statistics