An Estimation of Risky Module using SVM

SVM을 이용한 위험모듈 예측

  • 김영미 (한국원자력안전기술원 공학연구실) ;
  • 정충희 (한국원자력안전기술원 공학연구실) ;
  • 김현수 (충남대학교 전기정보통신공학부 컴퓨터)
  • Published : 2009.06.15

Abstract

Software used in safety-critical system must have high dependability. Software testing and V&V (Verification and Validation) activities are very important for assuring high software quality. If we can predict the risky modules of safety-critical software, we can focus testing activities and regulation activities more efficiently such as resource distribution. In this paper, we classified the estimated risk class which can be used for deep testing and V&V. We predicted the risk class for each module using support vector machines. We can consider that the modules classified to risk class 5 and 4 are more risky than others relatively. For all classification error rates, we expect that the results can be useful and practical for software testing, V&V, and activities for regulatory reviews.

안전-필수 분야에 사용되는 소프트웨어의 신뢰도(dependability)를 보장하기 위해 소프트웨어의 테스팅과 확인 및 검증활동이 매우 중요하다. 본 연구에서는 위험수준이 높은 소프트웨어 모듈을 소프트웨어 수명수기 초기에 예측하여, 테스팅과 확인 및 검증 활동에 대한 자원할당을 도울 수 있게 해준다. 다중 클래스 분류를 지원하는 SVM(Support Vector Machine)을 이용하여 소프트웨어 모듈의 잠재위험수준 을 예측한다 잠재위험수준이 상대적으로 높게 나온 모듈들에 대해 테스팅과 확인 및 검증을 집중적으로 실시함으로써 보다 효과적으로 소프트웨어의 품질을 향상시킬 수 있다. 또한, 원전의 계측제어계통에 사용되는 안전-필수 소프트웨어의 안전성 심사를 위한 대상 모듈을 샘플링할 때 활용할 수 있을 것으로 기대된다.

Keywords

References

  1. Boehm, B.W., Papaccio, P.N., Understanding and controlling software costs. IEEE Transcations on Software Engineering, Volume 14, Issue 10, pp. 1462-1477, 1988 https://doi.org/10.1109/32.6191
  2. Briand, L.T., Basili, V.R., Hetmanski, C., Deve-loping interpretable models for optimized set reduc-tion for identifying high-risk software components. IEEE Transactions on Software Engineering, Volume 19, Issue 11, pp. 1028-1034, 1993 https://doi.org/10.1109/32.256851
  3. Khoshgoftaar, TM., Munson, J.C., Prediction soft-ware development errors using complexity metrics. IEEE Journal on Selected Areas in Communi-cations, Volume 8, Issue 2, pp. 153-261, 1990 https://doi.org/10.1109/49.46879
  4. Porter, A., Selby, H., Empirically guided software development using metric-based classification trees. IEEE Software, Volume 7, Issue 2, pp. 46-54, 1990 https://doi.org/10.1109/52.50773
  5. Khoshgoftaar, T.M., Lanning, D,L., Pandya, A.S., A comparative study of pattern recognition tech-niques for quality evaluation of telecommunica-tions software, IEEE Journal on Selected Areas in Communications, Volume 12, Issue 2, pp. 208-217, 1994 https://doi.org/10.1109/49.272878
  6. Xing, F., Guo, P., Lyu, M.R., A novel method for early software quality prediction based on support vector machine, Proceedings of IEEE International Conference on Software Reliability Engineering, pp. 213-222, 2005 https://doi.org/10.1109/ISSRE.2005.6
  7. I. Gondra, Applying machine learning to software fault-proness prediction, The Journal of Systems and Software, Volume 81, Issue 2, pp. 186-195, 2008 https://doi.org/10.1016/j.jss.2007.05.035
  8. Karim O. Elish, Mahmoud O. Elish, Predicting defect-prone software modules using support vec-tor machines, The Journal of Systems and Software, Volume 81, Issue 5, pp. 649-660, 2008 https://doi.org/10.1016/j.jss.2007.07.040
  9. Burges. C., A Tutorial on Support Vector Machines for Pattern Recognition, Data Mining and Know-ledge Discovery, 1998 https://doi.org/10.1023/A:1009715923555
  10. T. Joachims, Making large-Scale SVM Learning Practical. Advances in Kernel Methods-Support Vector Learning, B. Scholkopf and C. Burges and A. Smola (ed.), MIT-Press, 1999
  11. Steve R, Gunn, Support Vector Machines for Cla-ssification and Regression, Technical Report, Uni-versity of Southampton, 10 May 1998
  12. http://svmlight.joachims.org/svm_multiclass.html