Statement-Level Bug Localization by Integrating Multiple Types of Features

doi:10.3778/j.issn.1002-8331.2411-0210

Abstract

Abstract: Existing deep learning-based statement-level bug localization methods perform bug localization by training deep learning models with different features, such as suspiciousness values computed by various bug localization techniques. These methods mainly focus on dynamic information from executing test cases and overlook the class imbalance issue between buggy and non-buggy statements, which limits the accuracy of bug localization. To address these issues, SMSDFL (combining spectrum, mutation, and semantic features for deep fault localization), a method that integrates multiple types of features for statement-level bug localization, is proposed. Using various spectrum-based and mutation-based bug localization methods to calculate the suspiciousness values, which are treated as spectrum and mutation features, and code statements are sliced by using data flow and control flow to obtain code fragments that are dependent on the corresponding statements, from which semantic features such as overlap and tightness metrics are calculated. Then generative adversarial network (GAN) is used to augment the feature data of bug statements to resolve the class imbalance issue between buggy and non-buggy statements. These three types of features are fused and input into the recurrent neural network to calculate the suspiciousness values of each statement, which are used to rank the statements in descending order for bug localization. To evaluate the performance of SMSDFL, experiments are conducted on the Defects4J dataset. The experimental results show that SMSDFL outperforms existing methods in bug localization performance, locates 166, 161, 151, and 68 more bug statements than Ochiai, MUSE, CGAN4FL, and GRACE, respectively, when only checking the first statement in the suspiciousness ranking.

Key words: bug localization, deep learning, neural network, generative adversarial network (GAN)

摘要： 现有基于深度学习的语句级缺陷定位方法通过深度学习模型学习各种传统缺陷定位方法计算出的可疑度值等特征来进行缺陷定位，其主要关注执行测试用例的动态信息，未关注缺陷语句与非缺陷语句之间的类不平衡问题，缺陷定位的准确性有待进一步提高。为此，提出了一种融合多类特征的语句级缺陷定位方法SMSDFL（combining spectrum，mutation and semantic features for deep fault localization），该方法将多种基于频谱和变异的缺陷定位方法计算的可疑度值作为频谱特征和变异特征，利用数据流和控制流对代码语句进行切片，获取与对应语句存在依赖关系的代码片段，以计算Overlap、Tightness等度量元作为其语义特征；采用生成对抗网络扩增缺陷语句的特征数据来解决缺陷语句与非缺陷语句之间的类不平衡问题；将这三类特征进行融合输入到递归神经网络中计算语句的可疑度值，并根据可疑度值对语句进行降序排序，以定位缺陷语句。为评估SMSDFL的缺陷定位性能在Defects4J数据集上进行了实验。实验结果表明，SMSDFL取得了比现有方法更好的缺陷定位效果，其中，仅检查可疑语句排名中的第一条语句时，SMSDFL比Ochiai、MUSE、CGAN4FL和GRACE分别多定位到166、161、151、68条缺陷语句。

关键词: 缺陷定位, 深度学习, 神经网络, 生成对抗网络

LI Jingwen, ZHANG Jiale, YUAN Zhihao, CUI Zhanqi. Statement-Level Bug Localization by Integrating Multiple Types of Features[J]. Computer Engineering and Applications, 2026, 62(3): 377-390.

李静雯, 张佳乐, 袁智豪, 崔展齐. 融合多类特征的语句级缺陷定位方法[J]. 计算机工程与应用, 2026, 62(3): 377-390.

References

[1] STRATEGIC P.The economic impacts of inadequate infrastructure for software testing[R].National Institute of Standards and Technology，2002.
[2] LI X，LI W，ZHANG Y Q，et al.DeepFL：integrating multiple fault diagnosis dimensions for deep fault localization[C]//Proceedings of the 28th ACM SIGSOFT International Symposium on Software Testing and Analysis.New York：ACM，2019：169-180.
[3] LI Y，WANG S H，NGUYEN T N.Fault localization to detect co-change fixing locations[C]//Proceedings of the 30th ACM Joint European Software Engineering Conference and Symposium on the Foundations of Software Engineering.New York：ACM，2022：659-671.
[4] YU K，LIN M X.Advances in automatic fault localization techniques[J].Chinese Journal of Computers，2011，34（8）：1411-1422.
[5] ZOU D M，LIANG J J，XIONG Y F，et al.An empirical study of fault localization families and their combinations[J].IEEE Transactions on Software Engineering，2021，47（2）：332-347.
[6] LI X，ZHANG L M.Transforming programs and tests in tandem for fault localization[C]//Proceedings of the ACM on Programming Languages，2017：1-30.
[7] SOHN J，YOO S.FLUCCS：using code and change metrics to improve fault localization[C]//Proceedings of the 26th ACM SIGSOFT International Symposium on Software Testing and Analysis.New York：ACM，2017：273-283.
[8] XUAN J F，MONPERRUS M.Learning to combine multiple ranking metrics for fault localization[C]//Proceedings of the 2014 IEEE International Conference on Software Maintenance and Evolution.Piscataway：IEEE，2014：191-200.
[9] LI Y，WANG S H，NGUYEN T.Fault localization with code coverage representation learning[C]//Proceedings of the 2021 IEEE/ACM 43rd International Conference on Software Engineering.Piscataway：IEEE，2021：661-673.
[10] QIAN J，JU X L，CHEN X.GNet4FL：effective fault localization via graph convolutional neural network[J].Automated Software Engineering，2023，30（2）：16.
[11] LEI Y，SUN C N，MAO X G，et al.How test suites impact fault localisation starting from the size[J].IET Software，2018，12（3）：190-205.
[12] TANTITHAMTHAVORN C，HASSAN A E，MATSUMOTO K.The impact of class rebalancing techniques on the performance and interpretation of defect prediction models[J].IEEE Transactions on Software Engineering，2020，46（11）：1200-1219.
[13] HAO D，PAN Y，ZHANG L，et al.A similarity-aware approach to testing based fault localization[C]//Proceedings of the 20th IEEE/ACM International Conference on Automated Software Engineering.New York：ACM，2005：291-294.
[14] YU Y B，JONES J，HARROLD M J.An empirical study of the effects of test-suite reduction on fault localization[C]//Proceedings of the 2008 ACM/IEEE 30th International Conference on Software Engineering.Piscataway：IEEE，2009：201-210.
[15] 张卓，雷晏，毛晓光，等.基于对抗生成网络的缺陷定位模型域数据增强方法[J].软件学报，2024，35（5）：2289-2306.
ZHANG Z，LEI Y，MAO X G，et al.Model-domain data augmentation using generative adversarial network for fault localization[J].Journal of Software，2024，35（5）：2289-2306.
[16] CHEN A R，CHEN T H，WANG S W.Pathidea：improving information retrieval-based bug localization by re-constructing execution paths using logs[J].IEEE Transactions on Software Engineering，2022，48（8）：2905-2919.
[17] DISTANTE D，FARALLI S.A two-phase bug localization approach based on multi-layer perceptrons and distributional features[C]//Proceedings of the 19th Conference on Computational Science and Its Applications.Cham：Springer，2019：518-532.
[18] WONG W E，GAO R Z，LI Y H，et al.A survey on software fault localization[J].IEEE Transactions on Software Engineering，2016，42（8）：707-740.
[19] ABREU R，ZOETEWEIJ P，VAN GEMUND A J C.An evaluation of similarity coefficients for software fault localization[C]//Proceedings of the 2006 12th Pacific Rim International Symposium on Dependable Computing.Piscataway：IEEE，2006：39-46.
[20] WONG W E，DEBROY V，LI Y H，et al.Software fault localization using DStar（D*）[C]//Proceedings of the 2012 IEEE Sixth International Conference on Software Security and Reliability.Piscataway：IEEE，2012：21-30.
[21] XIE X Y，CHEN T Y，KUO F C，et al.A theoretical analysis of the risk evaluation formulas for spectrum-based fault localization[J].ACM Transactions on Software Engineering and Methodology，2013，22（4）：1-40.
[22] PAPADAKIS M，LE TRAON Y.Metallaxis-FL：mutation-based fault localization[J].Software Testing，Verification and Reliability，2015，25（5/6/7）：605-628.
[23] HAN J X，HUANG C，SUN S Q，et al.bjXnet：an improved bug localization model based on code property graph and attention mechanism[J].Automated Software Engineering，2023，30（1）：12.
[24] YANG Y B，ZHOU Y M，LU H M，et al.Are slice-based cohesion metrics actually useful in effort-aware post-release fault-proneness prediction? an empirical study[J].IEEE Transactions on Software Engineering，2015，41（4）：331-357.
[25]EINARSSON A，NIELSEN J D.A survivor’s guide to Java program analysis with soot[EB/OL].（2008-07-17）[2024-09-20].https：//www.brics.dk/SootGuide/sootsurvivorsguide.pdf.
[26] OTT L M，THUSS J J.Slice based metrics for estimating cohesion[C]//Proceedings of the First International Software Metrics Symposium.Piscataway：IEEE，2002：71-81.
[27] 李晓卓，卿笃军，贺也平，等.基于源代码扩展信息的细粒度缺陷定位方法[J].软件学报，2022，33（11）：4008-4026.
LI X Z，QING D J，HE Y P，et al.Fine-grained bug location method based on source code extension information[J].Journal of Software，2022，33（11）：4008-4026.
[28] XIE H，LEI Y，YAN M，et al.A universal data augmentation approach for fault localization[C]//Proceedings of the 2022 IEEE/ACM 44th International Conference on Software Engineering.Piscataway：IEEE，2022：48-60.
[29] JOLLIFFE I T，CADIMA J.Principal component analysis：a review and recent developments[J].Philosophical Transactions Series A，Mathematical，Physical，and Engineering Sciences，2016，374（2065）：20150202.
[30] WANG S，LIU T Y，TAN L.Automatically learning semantic features for defect prediction[C]//Proceedings of the 2016 IEEE/ACM 38th International Conference on Software Engineering.Piscataway：IEEE，2017：297-308.
[31] GUO J，CHENG J H，CLELAND-HUANG J.Semantically enhanced software traceability using deep learning techniques[C]//Proceedings of the 2017 IEEE/ACM 39th International Conference on Software Engineering.Piscataway：IEEE，2017：3-14.
[32] ZENG M H，WU Y Q，YE Z T，et al.Fault localization via efficient probabilistic modeling of program semantics[C]//Proceedings of the 44th International Conference on Software Engineering.New York：ACM，2022：958-969.
[33] MENG X X，WANG X，ZHANG H Y，et al.Improving fault localization and program repair with deep semantic features and transferred knowledge[C]//Proceedings of the 2022 IEEE/ACM 44th International Conference on Software Engineering.Piscataway：IEEE，2022：1169-1180.
[34] ZHANG W，LI Z Q，WANG Q，et al.FineLocator：a novel approach to method-level fine-grained bug localization by query expansion[J].Information and Software Technology，2019，110：121-135.
[35] NI Z，BO L L，LI B，et al.An approach of method-level bug localization[J].IET Software，2022，16（4）：422-437.
[36] PARNIN C，ORSO A.Are automated debugging techniques actually helping programmers? [C]//Proceedings of the 2011 International Symposium on Software Testing and Analysis.New York：ACM，2011：199-209.
[37] CHAWLA N V，BOWYER K W，HALL L O，et al.SMOTE：synthetic minority over-sampling technique[J].Journal of Artificial Intelligence Research，2002，16：321-357.
[38] HE H B，BAI Y，GARCIA E A，et al.ADASYN：adaptive synthetic sampling approach for imbalanced learning[C]//Proceedings of the 2008 IEEE International Joint Conference on Neural Networks.Piscataway：IEEE，2008：1322-1328.
[39] TOMEK I.A generalization of the k-NN rule[J].IEEE Transactions on Systems，Man，and Cybernetics，1976，SMC-6（2）：121-126.
[40] DABLAIN D，KRAWCZYK B，CHAWLA N V.DeepSMOTE：fusing deep learning and SMOTE for imbalanced data[J].IEEE Transactions on Neural Networks and Learning Systems，2023，34（9）：6390-6404.
[41] ELYAN E，MORENO-GARCIA C F，JAYNE C.CDSMOTE：class decomposition and synthetic minority class oversampling technique for imbalanced-data classification[J].Neural Computing and Applications，2021，33（7）：2839-2851.
[42] KAUR H，PANNU H S，MALHI A K.A systematic review on imbalanced data challenges in machine learning：applications and solutions[J].ACM Computing Surveys，2019，52（4）：1-36.
[43] MOHAMMED R，RAWASHDEH J，ABDULLAH M.Machine learning with oversampling and undersampling techniques：overview study and experimental results[C]//Proceedings of the 2020 11th International Conference on Information and Communication Systems.Piscataway：IEEE，2020：243-248.
[44] WONG W E，QI Y.BP neural network-based effective fault localization[J].International Journal of Software Engineering and Knowledge Engineering，2009，19（4）：573-597.
[45] WONG W E，DEBROY V，GOLDEN R，et al.Effective software fault localization using an RBF neural network[J].IEEE Transactions on Reliability，2012，61（1）：149-169.
[46] ZHENG W，HU D S，WANG J.Fault localization analysis based on deep neural network[J].Mathematical Problems in Engineering，2016，2016（1）：1820454.
[47] ZHANG Z，LEI Y，MAO X G，et al.CNN-FL：an effective approach for localizing faults using convolutional neural networks[C]//Proceedings of the 2019 IEEE 26th International Conference on Software Analysis，Evolution and Reengineering.Piscataway：IEEE，2019：445-455.
[48] WONG E，WEI T T，QI Y，et al.A crosstab-based statistical method for effective fault localization[C]//Proceedings of the 2008 1st International Conference on Software Testing，Verification，and Validation.Piscataway：IEEE，2008：42-51.
[49] JAPKOWICZ N，STEPHEN S.The class imbalance problem：a systematic study[J].Intelligent Data Analysis，2002，6（5）：429-449.