Which Is More Important for Cross-Project Defect Prediction: Instance or Feature?

Yu, Qiao; Jiang, Shujuan; Qian, Junyan

doi:10.1109/sate.2016.22

Cited by 23 publications

(17 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Turhan et al [20] applied the neighbour filter method to remove those instances of the source project, whose features are not close enough to the ones of the target project. Besides, Yu et al [21] employed correlation-based feature selection to select features that have strong correlation with the target project. Ma et al [22] proposed a method called transfer Naive Bayes, using a data gravitation approach [23] to adjust the weights of training instances and build a naive Bayes classifier on them.…”

Section: Related Workmentioning

confidence: 99%

Software Defect Prediction via Attention-Based Recurrent Neural Network

Fan

Diao

et al. 2019

Scientific Programming

View full text Add to dashboard Cite

In order to improve software reliability, software defect prediction is applied to the process of software maintenance to identify potential bugs. Traditional methods of software defect prediction mainly focus on designing static code metrics, which are input into machine learning classifiers to predict defect probabilities of the code. However, the characteristics of these artificial metrics do not contain the syntactic structures and semantic information of programs. Such information is more significant than manual metrics and can provide a more accurate predictive model. In this paper, we propose a framework called defect prediction via attention-based recurrent neural network (DP-ARNN). More specifically, DP-ARNN first parses abstract syntax trees (ASTs) of programs and extracts them as vectors. Then it encodes vectors which are used as inputs of DP-ARNN by dictionary mapping and word embedding. After that, it can automatically learn syntactic and semantic features. Furthermore, it employs the attention mechanism to further generate significant features for accurate defect prediction. To validate our method, we choose seven open-source Java projects in Apache, using F1-measure and area under the curve (AUC) as evaluation criteria. The experimental results show that, in average, DP-ARNN improves the F1-measure by 14% and AUC by 7% compared with the state-of-the-art methods, respectively.

show abstract

Section: Related Workmentioning

confidence: 99%

Software Defect Prediction via Attention-Based Recurrent Neural Network

Fan

Diao

et al. 2019

Scientific Programming

View full text Add to dashboard Cite

show abstract

“…To share the historical data among different projects, researchers have proposed many cross-project defect prediction (CPDP) methods [8]- [10]. The CPDP methods aim to build the prediction model on one project (known as source project) with sufficient historical data, and use this model to predict on the other project (known as target project) [11].…”

Section: Introductionmentioning

confidence: 99%

“…In our previous work [11], we proposed the frameworks of instance filter and feature selection for CPDP, and we conducted preliminary experiments to compare their performance for CPDP. The results indicated that feature selection performed better than instance filter for CPDP.…”

Section: Introductionmentioning

confidence: 99%

An Empirical Study on the Effectiveness of Feature Selection for Cross-Project Defect Prediction

Qian

Jiang

et al. 2019

IEEE Access

Self Cite

View full text Add to dashboard Cite

Software defect prediction has attracted much attention of researchers in software engineering. At present, feature selection approaches have been introduced into software defect prediction, which can improve the performance of traditional defect prediction (known as within-project defect prediction, WPDP) effectively. However, the studies on feature selection are not sufficient for cross-project defect prediction (CPDP). In this paper, we use the feature subset selection and feature ranking approaches to explore the effectiveness of feature selection for CPDP. An empirical study is conducted on NASA and PROMISE datasets. The results show that both the feature subset selection and feature ranking approaches can improve the performance of CPDP. Therefore, we should select the representative feature subset or set a reasonable proportion of selected features to improve the performance of CPDP in future studies. INDEX TERMS Software defect prediction, cross-project defect prediction, feature selection, feature ranking.

show abstract

“…CPDP is twofold: instance and feature-focused works. Further, hybrid methods are also employed in CPDP (Xia et al, 2016;Yu et al, 2016). However, instance-focused works mainly include novel prediction models and there are few works that especially bring novelty in terms of data pre-processing (Li et al, 2017).…”

Section: Introductionmentioning

confidence: 99%

Is Complexity-based Clustering of Process Metrics as Effective as in Static Code Metrics

Öztürk¹

2019

BJMC

View full text Add to dashboard Cite

Defect prediction is not a new sub-field of software engineering. However, in this field, there are various research problems that are still attractive for many researchers. Cross-project defect prediction (CPDP), which is one of the popular issues of defect prediction, is still intriguing. To address this problem, generally instances or feature-focused experiments are performed but there is a lack of novel pre-processing methods. The main objective of this work is to propose a fuzzy clustering method that is based on complexity in CPDP. It helps selecting training data of CPDP. Hence, an opportunity that provides comparing static code and process metrics emerges. In this work, complexity-based fuzzy clustering that helps to select training instances of CPDP is proposed for process metrics. In the method, fuzzy membership levels are associated with a complexity value based on process metrics. In the experiment, 18 data sets including static code and process metrics together are employed. The findings of the experiment show that although static code metrics perform better than process metrics in terms of area under the curve (AUC), process metrics outperforms static code metrics in matthew's correlation coefficient (MCC) and F-measure parameters. Furthermore, in accordance with the used data sets, it has been detected that there is not any linear model among process metrics including number of revisions (NR), number of modified lines (NML), and number of distinct committers (NDC). This work asserts that the approach on the basis of training instance selection of CPDP yields remarkable success in process metrics. Moreover, in overall performance, process metrics are rather suitable for clustering-based instance selection.

show abstract

Which Is More Important for Cross-Project Defect Prediction: Instance or Feature?

Cited by 23 publications

References 29 publications

Software Defect Prediction via Attention-Based Recurrent Neural Network

Software Defect Prediction via Attention-Based Recurrent Neural Network

An Empirical Study on the Effectiveness of Feature Selection for Cross-Project Defect Prediction

Is Complexity-based Clustering of Process Metrics as Effective as in Static Code Metrics

Contact Info

Product

Resources

About