SOFTWARE QUALITY FORECASTING BASED ON REQUIREMENTS QUALITY ANALYSIS
Software quality assessment is an important component of its development in order to ensure both the latest and reasonable comparison of programs or their versions. Currently, software quality assessment is based on a number of international standards. At the same time, as a rule, methods of expert evaluation are used based on the metric approach of assessing the quality of software tools. Forecasting the quality characteristics of software tools is the next step in quality assurance, which should allow both obtaining reasonable achievable quality characteristics and serving, for example, for marketing research.
The paper examines existing approaches to predictive modeling of software quality characteristics. The factors that determine the forecast of software quality are analyzed. It was established that one of the main factors that determine the quality of the developed product is the quality of the requirements for it. A method of forecasting software quality characteristics based on requirements quality characteristics is proposed. Quality characteristics of requirements are established at the stage of testing (verification) of requirements. The initial data from which the predictive model is based are the quality characteristics obtained by the method of historical analogies. The predictive model assumes a simple linear dependence of the influence of the characteristics of the requirements on the predicted characteristics of the software tool, but at the same time, the number of requirements (the size of the project) and their classification are taken into account. In fact, the forecasting method consists in the established influence of the assessment of the quality characteristics of the requirements on the quality characteristics of the future project, and is carried out by the method of expert evaluation. The obtained values can be used both to substantiate the achievable quality characteristics and to improve "feedback" with customers to improve project requirements.
Lanubile F. and Visaggio G. Evaluating predictive quality models derived from software measures: lessons learned. Journal of Systems and Software. 1997. Vol. 38. P. 225–234.
Hong E., Wu C. Criticality models using SDL metrics set, 4 th Asia-Pacific Software Engineering and International Computer Science Conference. 1997. P. 23–30.
Evett M., Khoshgoftar T., Chien P. and Allen E. GP-based software quality prediction, Third Annual Genetic Programming Conference, 1998, P. 60–65.
Zhang D., Tsai J.J.P. Machine learning and software engineering, Software Quality Journal. 2003. Vol. 11, Issue 2. P. 87–119.
Ganesan K., Khoshgoftaar T., Allen E. Casedbased software quality prediction, International Journal of Software Engineering and Knowledge Engineering. 2000. Vol. 10. No 2. P. 139–152.
Khoshgoftaar T., Seliya N. Analogy-Based Practical Classification Rules for Software Quality Estimation, Empirical Software Engineering. 2003. Vol. 8, No 4. P. 325–350.
Khoshgoftaar T., Allen E.B., Deng J. Using regression trees to classify fault-prone software modules, IEEE Transactions on Reliability. 2002. Vol. 51, No 4. P. 455–462.
Netkachova K. I. Safety Case Metodology: Architecting principles. Радіоелектроніка і комп’ютерні системи. Харків: НАУ”ХАІ”. 2010. №7. С.109-112.
Hovorushchenko T.O. Teoretychni ta prykladni zasady informatsiynoyi tekhnolohiyi otsinyuvannya dostatnosti informatsiyi shchodo yakosti u spetsyfikatsiyakh vymoh do prohramnoho zabezpechennya – Kvalifikatsiyna naukova pratsya na pravakh rukopysu. Dysertatsiya na zdobuttya naukovoho stupenya doktora tekhnichnykh nauk za spetsialʹnistyu 05.13.06 «Informatsiyni tekhnolohiyi» (12 – Informatsiyni tekhnolohiyi) – Khmelʹnytsʹkyy natsionalʹnyy universytet, Ukrayinsʹka akademiya drukarstva, Lʹviv. 2018. 310 p. [in Ukrainian].
Biletsʹkyy T. P., Fedasyuk D. V. Prohnozuvannya defektiv u prohramnomu zabezpechenni alhorytmamy hlybynnoho navchannya CNNta RNN. Naukovyy visnyk NLTU Ukrayiny. 2021. T. 31, №2. S. 114 – 120. [in Ukrainian].
Subnis P., Kadam A. Software reliability growth model with bug ccle and duplicate detection techniques. Bharati Vidyapeech Deemed univ. College of Eng. Pune India. 2013. P 345-349.
Yakovyna V. S. Symetsʹ I. I. Prohnozuvannya defektiv prohramnoho zabezpechennya ansamblem neyronnykh merezh. Naukovyy visnyk NLTU Ukrayiny. 2021. T. 31, №6. P. 104 – 111. [in Ukrainian].
Naukovo-tekhnichni doslidzhennya u haluzi informatsiynykh tekhnolohiy: kolektyvna monohrafiya / zah. red. M. V. Kuzʹ. Akademiya tekhnichnykh nauk Ukrayiny. Ivano-Frankivsʹk: Vydavetsʹ Kushnir H.M. 2022. Vol 1. 152 p. [in Ukrainian].
Yerina A. M. Statystychne modelyuvannya ta prohnozuvannya. Kyyiv: KNTEU. 2001. 114 p. [in Ukrainian].
Nezamay B. S. Metodyka prohnozuvannya yakosti prohramnykh zasobiv, Perspektyvy rozvytku nauky, osvity ta tekhnolohiy v konteksti yevrointehratsiyi: zbirnyk tez dopovidey mizhnarodnoyi naukovo-praktychnoyi konferentsiyi (Poltava, 18 08 2022.), Poltava: TSFEND. 2022. P. 65-66. [in Ukrainian].
George Koelsch Herndon Requirements Writing for System Engineering. Virginia, USA. 2016. 409 p.
DSTU ISO/IEC 9126-1:2013. Prohramna inzheneriya: YAKISTʹ PRODUKTU Chastyna 1. Modelʹ yakosti (ISO/IEC TR 9126-1:2001, IDT). (Natsionalʹnyy standart Ukrayiny) [in Ukrainian].
Nezamay B. S. Prohnozuvannya yakosti prohramnoho zabezpechennya. 2022 International Conference on Innovative Solutions in Software Engineering (ICISSE) prysvyachena 50-richchyu vid dnya narodzhennya profesora Pavla Fedoruka. Ivano-Frankivʹk, 30 11 2022. Ivano-Frankivsʹk: PNU. 2022. p. 86-89. [in Ukrainian].