Cross-software defect prediction method based on adversarial discrimination

Abstract

The invention discloses a cross-software defect prediction method based on adversarial discrimination. The method comprises the following steps: selecting a source project and a target project; converting source codes in the source project and the target project into an abstract syntax tree, and extracting a node vector set; encoding the nodes, and converting the node vector set into an integer vector set; processing the integer vector set in the source project to train a source project feature extractor and a target project feature extractor at the same time, and extracting transferable codesemantic features in the source project and the target project; inputting the code semantic features capable of being migrated by the source project into a logistic regression classifier, training a cross-software defect prediction model, applying the defect prediction model to the target project, and performing defect prediction classification. According to the method, the adversarial discrimination method is used as one of powerful domain adaptive technologies, and the problem of feature distribution difference can be solved by minimizing the distance between the source project mapping distribution and the target project mapping distribution.

Description

technical field [0001] The invention relates to the field of software engineering, in particular to a cross-software defect prediction method based on adversarial discrimination. Background technique [0002] In the software development life cycle, if internal potential defects are discovered later, the later costs for repairing these defects will be greater. However, if each software module is fully and comprehensively tested, it is bound to inject too many human resources. So the project manager wants to identify possible defects in the software module in advance and focus on testing the module. Therefore, software defect prediction technology has attracted more and more attention from software engineering researchers and testers, and some software defect methods based on machine learning and deep learning have been proposed to detect defective files that may exist in software. [0003] The software defect prediction method based on machine learning uses features manuall...

AI Technical Summary

Problems solved by technology

However, there are some problems with this type of method, which does not consider the feature distribution difference between the source item and the target item, which also affects the defect prediction performance.

Images