A critical influence on software projects and on the success of their realization is exerted by issues related to the analysis and evaluation of the software requirements specifications (SRS). Today, when the number of high-budget software projects is growing rapidly, it is important to analyze the SRS, in particular, to determine their informativeness to assess the quality of software according to ISO 25010: 2011. The conducted analysis of the known tools for SRS' analysis showed that none of the known tools provides the possibility of determining the informativeness of the requirements specification. Therefore, it is necessary to design and implement the criteria and method for determining the informativeness of the SRS (which would be the theoretical basis for developing a future tool for determining the informativeness of requirements specifications), which is the purpose of this study. The paper proposes criteria and a method for determining the informativeness of the software requirements specifications, which determine the informativeness of the specification in terms of determining on its basis the characteristics of the software quality. The proposed criteria and method for determining the informativeness of the specifications of the software requirements allow determining the informativeness of the specification in terms of determining on its basis each of the 8 characteristics of the software quality.
The software development life cycle begins with the formation of a set of requirements and specification of software requirements based on them.
The largest number of software bugs are made at the stage of requirements formation [
82% of application rework is related to requirements bugs [
The importance and impact of business requirements on the success of the enterprise through
software projects found that 68% of companies suffer from the poor practice of technical
requirements and that these companies [
In fact, the study [
The risks of an insufficiently worked out stage of requirements formation are non-compliance with
project deadlines and financial overspending, which can lead to the closure of the project, and even
the collapse of the software company due to its financial instability [
In the requirements formation process, information losses emerge because of incompleteness and
difference of understanding of the users' needs and context of information – especially such losses are
significant for multidisciplinary software projects that are developed at the intersection of subject
areas [
Thus, the critical impact on software projects and the success of their implementation are issues related to the analysis and evaluation of the initial stages of the life cycle. Today, when the number of high-budget software projects is growing rapidly, it is important to analyze the specifications of software requirements, in particular, to determine their informativeness for assessing the quality of software according to ISO 25010: 2011.
Today, the key to assessing the quality of the software is the approach based on the SQuaRE model (ISO 25010:2011 standard). Software quality assessment according to ISO 25010:2011 [19] is as follows: based on the quality attributes specified in ISO 25023: 2016 [20], sub-characteristics and quality characteristics are evaluated.
Important criteria for specifying software requirements in terms of future software quality assessment are: 1. Completeness of the requirements specification 2. Sufficiency of information in the requirements 3. Informativeness of the requirements specification
Let's explore the known tools of the analysis of requirements in order to determine such criteria, as well as the possibility of their free use (Table 1).
Thus, as the conducted analysis of the known tools for software requirements specifications' analysis showed that none of the known tools provides the possibility of determining the informativeness of the requirements specification. Therefore, it is necessary to develop and implement criteria and method for determining the informativeness of software requirements specifications (which would be the theoretical basis for developing a future tool for determining the informativeness of requirements specifications), which is the purpose of this study.
Criteria for evaluating information, according to [30, 31], are: 1. Relevance – determining whether this information can help in the future to assess the quality of software according to ISO 25010 2. Veracity – the extent to which the description of the information is true; how the description of the information corresponds to reality 3. Significance – understanding of information, completeness of coverage of the subject of interest, timeliness of the information and its sufficiency for decision-making 4. Novelty – how the information's description is new for the user 5. Actuality – the degree of compliance of information with the current time 6. Objectivity – characterizes the independence of information from someone's opinion or consciousness, as well as from methods of obtaining 7. Completeness – information can be considered complete if it contains a minimum set of quality measures, but sufficient to make a correct decision 8. Compliance – the degree of compliance of information to the needs of consumers 9. Adequacy – compliance with the content of the actually obtained information and its expected content 10. Accessibility – a measure of the ability to obtain information
To assess the informativeness of the specification of software requirements, it is necessary to assess the informativeness of the quality measures of the specification in terms of their impact on the characteristics of the software quality.
Criteria of informativeness of quality measures of the SRS: 1. Relevance Rv = {Rsp, Abci}, where Rsp € [0;1] – the rank of the influence of the quality measure of the specification on a certain characteristic, where 0 – corresponds to the smallest influence of the quality measure of the specification on the characteristic, 1 – corresponds to the greatest influence; Abci € [0;1] – the rank of the ability of the quality measure to clarify the value of the software quality characteristic, where 0 means that the measure least clarifies the value of the characteristic, 1 – means the greatest degree of clarification 2. Veracity Rb € [0;1], where 1 – it's possible to trust to the measure of determining the characteristic of software quality; 0 – cannot be trusted in principle 3. Significance Sf = {Ust, Cnc}, where Ust € [0;1] – the rank of understanding the quality measure, where 0 – the measure is unclear, 1 – the measure is clear; Cnc € [0;1] – the rank of completeness of the coverage of the quality characteristic by the measure, where 0 – the measure least covers the characteristic, 1 – the most illuminates the characteristic 4. Objectivity Obj € [0;1], where 0 – quality measure depends on someone's opinion or consciousness, as well as on the methods of obtaining, 1 – independent measure 5. Compliance Qi € [0;1], where 0 – the lowest degree of conformity of the quality measure to the characteristic, 1 – the highest degree
As for other criteria for evaluating information, they will not be needed to assess the informativeness of the software specification due to their redundancy and repeatability.
Taking into account the above criteria of informativeness of the measures of the specification of requirements, the method of determining the informativeness of the specification of software requirements consists of the following stages: 1. Division of the interval into subintervals for each software quality characteristic quality
Evaluation of all measures of each quality characteristic according to each of the criteria of informativeness of quality measures of the specification of requirements measure of the corresponding rank
Sorting of quality measures of the specification of requirements and assignment to each Formation of a matrix of informativeness of measures for each characteristic of the software Calculation of the indicator of informativeness of the specification on the definition of each of the software quality characteristics
The interval [0; 1] is divided into subintervals depending on how many quality measures of the specification affect the quality characteristic: − 1.
(1)
Then for an ideal specification, in which there are absolutely all the measures on which the quality characteristics depend, the number of subintervals, the step and the division of the interval into subintervals for each software quality characteristic are presented in Table 2.
subintervals
corresponding rank. characteristic (j=1..8):
Ij
=
Then there is an expert assessment of the importance of each measure of quality characteristic according to a certain criterion, sorting of measures is performed, and each measure is assigned a The result of such assessments is the matrix of informativeness of measures for each j-th Rsp1j
…
Qi1j … Qinj 9 23 44 25 8 15 26 16 8 22 43 24 7 14 25 15
1/8=0.125 1/22=0.0455 1/43=0.0233 1/24=0.0417 1/7=0.1429 1/14=0.0714
1/25=0.04 1/15=0.0667
Dividing the interval into subintervals (to determine the ranks) [0; 0.125;…;
0.875; 1] [0; 0.0455;…;
0.9545; 1] [0; 0.0233;…;
0.9767; 1] [0; 0.0417;…;
0.9583; 1] [0; 0.1429;…;
0.8571; 1] [0; 0.0714;…; 0.9286; 1] [0; 0.04;…;
0.96; 1] [0; 0.0667;…; 0.9333; 1] where the first line contains estimates for the measure №1 of j-th characteristic, the second line contains estimates for the measure №2 of j-th characteristic, etc., the last line contains estimates for the measure №n of j-th characteristic (number of measures are different for each characteristic and maximum of the number of measures are presented in Table 2 for each characteristic). Thus, we obtain 8 matrices of informativeness of measures for 8 main characteristics of software quality.
Next, it is necessary to assess the informativeness of the specification for the definition of each of the characteristics. To do this, it's necessary to calculate for each matrix the number of elements belonging to a certain range – Qiij, after this to find the informativeness of the specification for determining each of the characteristics as the ratio of the obtained number Qiij to the total size of each j-th matrix (7*nj): characteristic of the software quality.
Thus, the criteria and method for determining the informativeness of the specifications of the software requirements are proposed, which determine the informativeness of the specification in terms of determining on its basis the characteristics of the software quality.
ISO 25010, ISO 25023):
Number of Functions Functional Implementation Completeness Functional Adequacy Functional Implementation Coverage Operation Time Number of Inaccurate Computations Encountered by Users Number of Data Items Computational Accuracy
Precision For example, let's consider the measures on which Functional Suitability depends (according to The number of measures on which the Functional Suitability characteristic depends is 9. All 9 measures of the Functional Suitability are present in the first considered specification. Then the number of subintervals is 8, so the step is 1/8 = 0.125. We then divide the interval as follows: [0; 0.125; …; 0.875; 1]. Next, we evaluate with the help of experts each measure of the characteristic according to a certain criterion, sort the measures and assign each measure the appropriate rank.
For example, let's consider how the experts recommended assessing the influence of each quality measures of the specification on the characteristic Functional Suitability. Thus, the most influential, according to experts, is the measure Number of Functions (rank 1), next in influence is the measure Operation Time (rank 0.875), then the measure Functional Implementation Completeness (rank 0.75), then Functional Implementation Coverage (rank 0.625), then Functional Adequacy (rank 0.5), then Precision (rank 0.375), then Number of Data Items (rank 0.25), then Computational Accuracy (rank 0.125), and the least influential is the Number of Inaccurate Computations Encountered by Users (rank 0).
Similarly, the experts recommended assessing the rank of the ability of each quality measure to clarify the value of Functional Suitability, the veracity of each measure to determine the characteristic Functional Suitability, the rank of understanding of each quality measure of Functional Suitability, the rank of completeness of the coverage of Functional Suitability by each measure, the objectivity of each measure of the Functional Suitability characteristic, compliance of each measure concerning the definition of the Functional Suitability characteristic.
The result of such estimates is a matrix of informativeness of measures for the Functional Suitability characteristic:
1 0.75 0.5 0.625 0.875
0 0.25 0.125 0.375
Let's calculate for the obtained matrix the number of elements that belong to a certain range – Qiij. For example, let's calculate the number of elements that belong to the range [0.75; 1], because experts recommend just such an interval for Functional Suitability, if Functional Suitability depends on not all 9 measures, but only 5 measures. Under such conditions, Qiij = 14 for the obtained matrix.
Then the informativeness of the specification for determining the Functional Suitability characteristic is:
14 _ 2 =
∗ 100% = 40%.
7 ∗ 5
It is obvious that if the specification of requirements contains a smaller number of quality measures for a certain characteristic, then the informativeness of the specification for determining such a characteristic is lower (in the considered examples – by 15.56% while reducing the number of measures by 4). Similarly, the matrices of the informativeness of the remaining 7 software quality characteristics can be obtained and, accordingly, the informativeness of the specification for determining each of the remaining 7 software quality characteristics can be calculated.
Therefore, the proposed criteria and method for determining the informativeness of the specifications of the software requirements allow determining the informativeness of the specification in terms of determining on its basis each of the 8 characteristics of software quality.
A critical influence on software projects and on the success of their realization is exerted by issues related to the analysis and evaluation of the software requirements specifications (SRS). Today, when the number of high-budget software projects is growing rapidly, it is important to analyze the SRS, in particular, to determine their informativeness to assess the quality of software according to ISO 25010.
The conducted analysis of the known tools for SRS’ analysis showed that none of the known tools provides the possibility of determining the informativeness of the requirements specification. Therefore, it is necessary to design and implement the criteria and method for determining the informativeness of the SRS (which would be the theoretical basis for developing a future tool for determining the informativeness of requirements specifications), which is the purpose of this study.
The paper proposes criteria and a method for determining the informativeness of the software requirements specifications, which determine the informativeness of the specification in terms of determining on its basis the characteristics of the software quality. The proposed criteria and method allow determining the informativeness of the specification in terms of determining on its basis each of the 8 characteristics of the software quality.
[13] A. Melnyk, B. Dunets, FFT processor IP Cores synthesis on the base of configurable pipeline architecture, in: Proceedings of the 7th International Conference on Experience of Designing and Application of CAD Systems in Microelectronics, CADSM’2003, Slavske, 2003, pp. 211–213. doi: 10.1109/CADSM.2003.1255034. [14] T. Hovorushchenko, A. Herts, Ye. Hnatchuk, Concept of intelligent decision support system in the legal regulation of the surrogate motherhood. CEUR-WS 2488 (2019) 57-68. [15] O. Drozd, M. Kuznietsov, O. Martynyuk, M. Drozd, A method of the hidden faults elimination in FPGA projects for the critical applications, in: Proceedings of 2018 IEEE 9th International Conference on Dependable Systems, Services and Technologies, DESSERT’2018, Kyiv, 2018, pp. 231 – 234. doi: 10.1109/DESSERT.2018.8409131. [16] A. Drozd, J. Drozd, S. Antoshchuk, V. Nikul, M. Al-dhabi, Objects and methods of on-line testing: main requirements and perspectives of development, in: Proceedings of 2016 IEEE EastWest Design & Test Symposium, EWDTS’2016, Yerevan, 2016, pp. 72 – 76. doi: 10.1109/EWDTS.2016.7807750. [17] M. Drozd, A. Drozd, Safety-related instrumentation and control systems and a problem of the hidden faults, in: Proceedings of the 10th International Conference on Digital Technologies, Zhilina, 2014, pp. 137–140. doi: 10.1109/DT.2014.6868692. [18] A. Drozd, M. Drozd, V. Antonyuk, Features of hidden fault detection in pipeline components of safety-related system. CEUR-WS 1356 (2015) 476–485. [19] ISO/IEC 25010:2011. Systems and software engineering. Systems and software Quality Requirements and Evaluation (SQuaRE). System and software quality models. [Introduced 01.03.2011]. Geneva (Switzerland), 2011. (International standard). [20] ISO 25023:2016. Systems and software engineering. Systems and software Quality Requirements and Evaluation (SQuaRE). Measurement of system and software product quality. [Introduced 31.03.2016]. Geneva (Switzerland), 2016. (International standard). [21] Software Requirements Engineering Tools, 2018. URL: http://ecomputernotes.com/softwareengineering/softwarerequirementsengineeringtools. [22] 30 Best Requirements Management Tools in 2019, 2019, URL: https://www.guru99.com/requirement-management-tools.html. [23] K. Verma, A. Kass, Requirements analysis tool: a tool for automatically analyzing software requirements documents. Lecture Notes in Computer Science 5318 (2008) 751-763. doi: 10.1007/978-3-540-88564-1_48. [24] H. Mahmood, M. Rehman, Tools for software engineers. IMPACT: International Journal of
Research in Engineering & Technology 3 10 (2015) 75-86. [25] V. Jones, J. Murray, Evaluation of current requirements analysis tools capabilities for IV&V in the requirements analysis phase, 2017, URL: https://www.slideserve.com/shlomo/evaluation-ofcurrent-requirements-analysis-tools-capabilities-for-ivv-in-the-requirements-analysis-phase. [26] D. Raffo, R. Ferguson, Evaluating the Impact of The QuARS Requirements Analysis Tool Using Simulation, 2018, URL: https://pdfs.semanticscholar.org/7e7d/4e6f5ab13d00ca57c87711e30cd080730f34.pdf. [27] F. Konig, L. Ballejos, M. Ale, A semi-automatic verification tool for software requirements specification documents, in: Proceedings of Simposio Argentino de Ingeniería de Software.
Argentina, 2017. [28] O. Pomorova, T. Hovorushchenko, Research of artificial neural network's component of software quality evaluation and prediction method, in: Proceedings of the 2011 IEEE 6-th International Conference on Intelligent Data Acquisition and Advanced Computing Systems: Technology and Applications, IDAACS’2011, Prague, 2011, vol. 2, pp. 959-962. doi: 10.1109/IDAACS.2011.6072916 [29] O. Pomorova, T. Hovorushchenko, Artificial neural network for software quality evaluation based on the metric analysis, in: Proceedings of IEEE East-West Design & Test Symposium, EWDTS’2013, Kharkiv, 2013, pp.200-203. doi: 10.1109/EWDTS.2013.6673193 [30] I. Nezhdanov, Information assessment criteria (importance, accuracy, significance), 2010. URL: http://www.police-russia.ru/showthread.php?t=44683. [31] Information bases of computing: Information, 2015. URL: https://intuit.ru/studies/courses/3657/899/info.