Розроблення гібридного методу обчислення складності програмного забезпечення
The use of code metrics allows software developers and project managers to evaluate various features of the software (to be built or already in existence), predict workload, determine software complexity and reliability, and quantify the quality of software systems being developed. Articles written...
Saved in:
| Date: | 2022 |
|---|---|
| Main Authors: | , |
| Format: | Article |
| Language: | English |
| Published: |
The National Technical University of Ukraine "Igor Sikorsky Kyiv Polytechnic Institute"
2022
|
| Subjects: | |
| Online Access: | https://journal.iasa.kpi.ua/article/view/240177 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Journal Title: | System research and information technologies |
| Download file: |  |
Institution
System research and information technologies| _version_ | 1866302768384311296 |
|---|---|
| author | Kazimov, Tofig Bayramova, Tamilla |
| author_facet | Kazimov, Tofig Bayramova, Tamilla |
| author_sort | Kazimov, Tofig |
| baseUrl_str | http://journal.iasa.kpi.ua/oai |
| collection | OJS |
| datestamp_date | 2022-10-17T22:12:39Z |
| description | The use of code metrics allows software developers and project managers to evaluate various features of the software (to be built or already in existence), predict workload, determine software complexity and reliability, and quantify the quality of software systems being developed. Articles written in recent years have proposed various methods for solving this problem. However, there is still no very effective approach to measuring software complexity. This article provides a brief overview of existing software complexity metrics and proposes a new hybrid method for computing software complexity. The proposed hybrid method for evaluating software complexity combines the key features of the Halsted, Maccabe, and SLOC metrics and also allows for a more efficient assessment of complexity. |
| doi_str_mv | 10.20535/SRIT.2308-8893.2022.2.02 |
| first_indexed | 2025-07-17T10:27:30Z |
| format | Article |
| fulltext |
T.H. Kazimov, T.A. Bayramova, 2022
32 ISSN 1681–6048 System Research & Information Technologies, 2022, № 2
TIДC
ПРОГРЕСИВНІ ІНФОРМАЦІЙНІ ТЕХНОЛОГІЇ,
ВИСОКОПРОДУКТИВНІ КОМП’ЮТЕРНІ
СИСТЕМИ
UDC 3338.01
DOI: 10.20535/SRIT.2308-8893.2022.2.02
DEVELOPMENT OF A HYBRID METHOD FOR CALCULATION
OF SOFTWARE COMPLEXITY
T.H. KAZIMOV, T.A. BAYRAMOVA
Abstract. The use of code metrics allows software developers and project managers
to evaluate various features of the software (to be built or already in existence),
predict workload, determine software complexity and reliability, and quantify the
quality of software systems being developed. Articles written in recent years have
proposed various methods for solving this problem. However, there is still no very
effective approach to measuring software complexity. This article provides a brief
overview of existing software complexity metrics and proposes a new hybrid
method for computing software complexity. The proposed hybrid method for
evaluating software complexity combines the key features of the Halsted, Maccabe,
and SLOC metrics and also allows for a more efficient assessment of complexity.
Keywords: software engineering, software complexity, complexity metrics, hybrid
method.
INTRODUCTION
One of the most dynamically developing spheres of modern life is information
technology. Today, people’s lives are organized in such a way that even people
who are far from information technology use them to achieve their goals. The
introduction of automated information systems not only reduces the number of
operations performed by a person, but also creates new problems. As software
grows in size and complexity, the number of bugs increases. Software projects
that at first glance appear to be successful may be stopped due to such errors, or
the code may be rewritten. This leads to a slight increase in budget expenditures.
The world’s leading software companies are working on these problems. Various
standards organizations (ISO/IEC, IEEE) have developed hundreds of standards
covering all stages of software life to improve its quality. Software quality is a
complex and multifaceted concept. After the advent of computers in the defense
industry in the 1970s, metrics began to emerge to evaluate software performance
and quality management.
Software measurement is one of the most important issues in developing
quality software. In the words of Tom DeMarco [1], “you can’t control what you
can’t measure”.
At the moment, a large number of different software metrics have been
developed. Software metrics fall into three categories (Fig.1):
Development of a hybrid method for calculation of software complexity
Системні дослідження та інформаційні технології, 2022, № 2 33
Product metrics are product characteristics such as performance, design,
size, quality level, and complexity.
Process Metrics. These metrics are used to improve software
development and maintenance processes.
Project Metrics. These metrics describe the performance and
characteristics of the project.
Among these indicators, software complexity indicators are of particular
importance. In the 10th edition of
Sommerville’s book Software Engineering,
he noted that one of the most important
tasks in the development of the modern
software industry is the management of
software complexity [4]. Research shows
that when the project size approaches 5
million lines, the number of defects begins
to increase dramatically (Fig. 2). This can
be explained by the significant
complication of large-scale projects [5].
Complexity is the degree where it is difficult to verify and understand the
design or application of a system or component. The problem of software
complexity estimation is widely studied in the field of software engineering. The
complexity of the software may vary depending on the choice of algorithm,
design, choice of programming language, writing code. Research shows that
increasing complexity increases the number of bugs in software, which makes it
difficult (in some cases impossible) to maintain and improve programs and makes
it difficult to test certain modules. The legibility of the program code directly
depends on the level of complexity of the program. The relationship between
indicators of software complexity and various attributes of a software system is
shown in Fig. 3.
The importance of program complexity indicators can be illustrated as
follows [7]:
Difficulty metrics can help people predict and sustain projects.
Complexity metrics can help estimate the amount of programming and
development costs, and estimate maintenance costs.
Software metrics
Fig. 1. Software metrics
Fig. 2. Dependence of KLOC
T.H. Kazimov, T.A. Bayramova
ISSN 1681–6048 System Research & Information Technologies, 2022, № 2 34
Difficulty scores can help you choose the most suitable program with the
same functionality. The less complex the programs, the better they are.
Complexity measures can be used to predict defects or errors [8].
By determining the complexity of software systems, the overall workload
and personal contribution of each person working on software modules to the
overall work can be assessed in evaluating the performance of software
developers [9].
RELEATED WORK
Application of program code metrics allows professionals who work on the
project to evaluate various feauters of existing or to be created software, to predict
the scope of work, quantitatively characterize these or other project solutions, to
evaluate quality of prepared systems, complexity and realiability of software
[9–13].
The large amount of data and functionality required in today’s enterprise
systems presents many challenges for software developers. It is difficult to
maintain a balance between software complexity and ease of use, which are
indicators of the quality of software, as the complexity of the software increases
and its use becomes more difficult. To ensure the quality of the software and the
high level of project management, it is necessary to control the complexity and
other related characteristics. Large companies such as SAP and Oracle are
currently losing market share due to the cost and complexity of the software
product [14].
Numerous studies show that complexity leads to an increase in the number
of vulnerabilities in software. Programmers face a number of challenges when
trying to make changes to complex software components. Software developers
should strive for minimal complexity, as increasing complexity creates security
risks that not only damage the business, but also damage its reputation. Violation
of safety measures can also pose a serious threat to human health. Some experts
deliberately complicate the same program so that they do not write it twice. It is
necessary to simplify the management of software security. The authors
substantiate the relationship between security and software complexity (Fig. 4) [15].
Different software development organizations use different metrics to
measure and maintain the quality of software code. In [16, 17], the authors
Fig. 3. Software complexity and various attributes of a software
Development of a hybrid method for calculation of software complexity
Системні дослідження та інформаційні технології, 2022, № 2 35
conducted comparative analytical studies using Halstead metrics and proved that
software complexity metrics can be used to measure various characteristics. The
authors analyzed the program code written for the same function in the
programming languages Python, C, JavaScript, and Java. The complexity of these
program codes was calculated based on the Halstead scores. Experience has
shown that python is more convenient and simpler, while Java is a modern,
powerful, but complex programming language. They showed that program
performance is important throughout the life of a project, and they concluded that
early calculations have a direct impact on fixing errors and thus saving time and
budget.
In [18], the authors point out that the main reason for software vulnerabilities
is its complexity. Failure to discover software vulnerabilities in a timely manner
can jeopardize confidentiality, completeness, and availability. Studies of 12
programs, differing in characteristics and sizes, have shown that the complexity of
all software components (size, structure, etc.) is important for predicting
vulnerabilities.
In their articles, Shin and William examined code complexity metrics that
can be used to predict software vulnerabilities, and concluded that the more
complex the software code, the more vulnerabilities. Their results showed a
correlation between difficulty scores and vulnerabilities in the Mozilla JavaScript
Engine [19].
In [20], the authors investigated the influence of complexity, coupling and
cohesion on the number of vulnerabilities in a program. Experiments with Mozilla
Firefox have empirically proven that there is a relationship between these parame-
ters and vulnerabilities in the program. Vulnerabilities cannot be discovered in the
early stages of software development. Metrics such as complexity, coupling and
cohesion can be calculated early. Calculating these metrics can help software de-
velopers identify potential vulnerabilities.
In [21], the authors presented a new approach to measuring the quality of
software using fuzzy metrics obtained as a result of software design. This metric
begins with an assessment of the difficulty rating for each class, which is itself
assessed based on the difficulty rating of the class attributes and the difficulty
rating of the class methods. Then, to assess the quality metrics of the fuzzy code,
Fig. 4. Security and software complexity
T.H. Kazimov, T.A. Bayramova
ISSN 1681–6048 System Research & Information Technologies, 2022, № 2 36
they followed a pattern to explain the relationship between class complexity and
LOC for that class, and between method complexity and LOC for that method.
The paper [23] examines the problem of software complexity and its impact
on software errors. Having studied various studies, the authors came to the
conclusion that with the increase in the complexity of the software system, the
threats to its security also increase. While most studies have shown that bugs in
software systems are related to their complexity, some researchers have noted that
the relationship between complexity and the occurrence of bugs is weak [24].
The software development process, including documentation, design,
software, testing, and support, can be measured statistically. Thus, the quality of
the software can be effectively controlled. Software metrics are very important in
software engineering research [25] provides a summary of software metrics and
their types. The authors note that measuring the complexity of software is an
integral part of software performance and affects the price and reliability of
software products.
Software users assume there are no bugs in the program, but software
developers know that it is very difficult, and in some cases impossible, to write
program code without bugs. This complexity is mainly due to the intrinsic
complexity of the program and the problems that arise when developing and
testing the program. In [26], the authors demonstrated the relationship between
the complexity of software and its reliability in each specific case.
In [27], the authors analyzed aspects influencing the complexity of the
program based on various metrics. A large amount of program code to some
extent affects the programmer’s thought process and leads to more errors in the
program. At the same time, if the number of modules and branches in the program
is large, the volume of information exchange is large, and the program is very
complex, this will cause problems for the program testers and reduce the quality
of testing.
Software complexity plays an important role in reducing the effort required
to build and maintain software, and in improving testing efficiency and software
quality. The more complex the solution of the program, the more errors it creates
[28] examines four software metrics, their importance, strengths and weaknesses.
The authors come to the conclusion that each method covers a part and takes into
account a certain group of parameters. Therefore, it is important to use a
combination of these metrics to measure software complexity.
HYBRID SOFTWARE COMPLEXITY CALCULATION METHOD
The hybrid software complexity method presented in this study combines the
advantages of the SLOC, Halstead, Maccab metrics and the module connectivity
metric.
Number of lines of code (LOC). When assessing the complexity of a
software product, three groups of indicators are usually used: indicators that
determine the size of the program, indicators of the complexity of the program
flow, and complexity of the data flow of the program. The first group is more
common because the metrics are simple. The program size traditionally means the
number of source lines of the program (SLOC – Source Lines Of Code) [29].
Development of a hybrid method for calculation of software complexity
Системні дослідження та інформаційні технології, 2022, № 2 37
Lines of Code (LOC) or Source Lines of Code (SLOC) are used to measure the
size of a program by counting the number of lines in the text of the program’s
source code. LOC measures the amount of code, can be used to compare or
evaluate programs that use the same programming language and are coded using
the same coding standards. Lines of code are completely dependent on the
programming language and may differ during the conversion of the program code
to any other programming language. This indicator cannot be considered effective
in assessing complexity.
Halsted metric. M.H. Halsted was the first researcher to write a
mathematical formulation of software metrics [30]. Table 1 shows the main
Halsted metrics.
T a b l e 1 . Halsted metrics
Program dictionary 21 hhh
Program length 21 NNN
Program scope hNV 2log
Program complexity )()2( 221 hNhD
Implementation effort or program comprehensibility DVE
Expected number of errors in the program 3000/667,0EB
The Halsted metric is based on four measurable program characteristics:
1h — the number of unique program operators;
2h — the number of unique software operands;
1N — total number of program operators;
2N — total number of program operands.
McСabes cyclomatic complexity. McCabe proposed this metric in 1976.
McCabe’s cyclomatic complexity metric is one of the most common indicators
for evaluating software complexity. This indicator is calculated based on the
control-flow graph of the program and is an indicator of the complexity of the
control structures.
The formula for calculating cyclomatic complexity is as follows:
pneC 2 ,
where e — the number of edges of the graph; n — the number of nodes of the
graph; p — the number of connected components.
This method takes into account all kinds of cyclic and conditional operators,
as well as the complexity of their logical predicates. Linear operators are ignored
here. Despite the passage of 45 years, it remains relevant because it more
accurately expresses the complexity of the fundamental structures of software.
The author of the metric notes the high correlation of the metric with errors.
Therefore, the use of this metric to evaluate errors is reasonable and logical.
Cyclomatic complexity allows not only to estimate the labor costs and the cost of
software projects, but also to make the necessary management decisions taking
into account risks [32, 33].
T.H. Kazimov, T.A. Bayramova
ISSN 1681–6048 System Research & Information Technologies, 2022, № 2 38
The program module coupling metric distinguishes between data
communication, data structure, control, common area (global data), content, as
well as external communication, message communication, subclass
communication, time communication, and no communication. The advantages of
these metrics are that they were developed back in the 80s, are well studied, are
often used, and the calculation of values for metrics is automated. These metrics
are intended for structured programming, but can be applied to object-oriented
programming.
A hybrid method for calculating the complexity of a program code. The
proposed method provides an integrated approach to the complexity of the
software system and takes into account the following disavantages of the McCabe
and Halsted metrics and LOC:
it cannot be argued that a program with the maximum number of lines of
code is more complex, sometimes a small program can be quite complex;
linear, cyclic, conditional operators in the system of Halsted metrics have
the same complexity, which is not true;
all cyclic, conditional operators in the McCabe metrics system have the
same complexity, which is not true. It is assumed that the more nested cyclic and
conditional (multiple branching) statements, the more complex the program code
becomes and the greater the likelihood that the programmer can make a logical
error;
models do not take into account the inter-module structural complexity of
the software, which generates a significant number of defects.
In this method, the main parameters affecting the complexity are the
algorithmic complexity of each program module, the number of standard called
programs ready for use, and the total number of intermodule links.
When solving problems algorithmically, it often becomes necessary to create
a cycle containing another cycle in its body. Such loops are called nested loops.
Sometimes you have to check several conditions in a row. To solve the problem
of redundancy, you can nest conditional statements inside each other. This is
called multiple branching of conditional statements. Nested loops and multiple
branches in the code complicate it. At the same time, the number of interacting
modules and standard software applications available in these modules directly
affects the complexity of the program.
The complexity of the program code can be calculated using the formula
(higher value is considered better):
rnNEC )(loglog 22 ,
where N — the number of all operators in the program code; n — total number
of conditional and cyclic operators; r — the number of intermodular couplings;
E — parameter depending on the number of cyclic and conditional operators in
the program code. This parameter is calculated using the following formula:
n
i
imiE
0
)1( , ),...,1,0( ni ,
Development of a hybrid method for calculation of software complexity
Системні дослідження та інформаційні технології, 2022, № 2 39
im is the total number of conditional and cyclic operators in which conditional
and cyclic statements of number i are nested. (For example, if there are 2 cyclic
statements in the program code and three nested cyclic statements in each of
them, then 3i , 23 m ).
Let’s calculate the complexity of two programs written in C++ (Fig. 5
and 6). First, we received an expert assessment of the complexity of these
programs based on the assessment of 3 experts on a 10-point scale. All experts
rated the algorithm in Fig. 6 as complex, since this algorithm has more nested
cyclic and conditional operators. The calculation results are shown in Table 2.
T a b l e 2 . Software complexity metrics
Complexity
metrics
Calculation Program 1 Program 2
LOC Number of lines of code 27 30
Expert
assessment
3 4
MacCabe
For a single program
2 neC
e=15; n=12 6 e=17, n=13 6
1 #include<iostream>
2 using namespace std;
3 int main {}
4 {
5 int n=3;
6 int a[n][n];
7 int m=0
8 for {int i=1; i<=n;i++}
9 for {int j=1; j<=n;j++}{
10 cin>>a[i][j];
11 if (a[i][j]==i+j)
12 m=m+1;
13 else
14 m=m;
15 }
16
17 if(m>0)
18 for {int i=1; i<=n;i++}
19 for {int j=1; j<=n;j++}{
20
21 if (a[i][j]==i+j)
22 cout<<a[i][j];
23 else
24 continue;
25 }
26 else
27 cout<<”No such element”;
28 return 0;
29 }
m:=0
Fo
For i, j=1 to N m:=m
m>0
m:=m+1
A(i,j)
A(i,j)=i+j
A(i,j)=i+j
A(i,j)
Fig. 5. Finding complexity of program 1
T.H. Kazimov, T.A. Bayramova
ISSN 1681–6048 System Research & Information Technologies, 2022, № 2 40
Continued Tabl. 2
Complexity
metrics Calculation Program 1 Program 2
Hybrid
n
i
imiE
0
)1(
rnNEC )(loglog 22
N=108, \
n=5, r=0,
m0=3 (i=0)
m1=2 (i=1),
9,49
N=101,
n=5, r=0,
m0=3 (i=0),
m1=0 (i=1),
m2=1 (i=2),
m3=1 (i=3)
9,58
As can be seen from Table 2 according to McCabe, the complexity of pro-
grams is the same, because in this method all conditional and cyclic operators
have the same level of complexity. The hybrid method takes into account the
nesting of cycles, so the result is closer to the expert assessment.
In another example, let’s calculate the complexity of a program code written
for the same function in C++ and Java using the metrics mentioned above (Fig. 7
and 8). Depending on the chosen programming language, the complexity may
increase. Calculating the complexity will help you choose the simpler of the two
programs. The calculation results are shown in Table 3.
Fig. 6. Finding complexity of program 2
Start
N, Z
N>1
i=1, N–1
V=Zi, L=i
j=j+1, N
Zj<V
V=Zj, L=j
L=i
Z
End
ZL=Zi, Zi=V
1 #include<iostream>
2 using namespace std;
3 int main {}
4 {
5 int n;
6 cin>>n;
7 int z[n];
8 for (int i=1; i<=n;i++){
9 cin>>z[i];
10 }
11 if (n>1)
12 for (int i=1; i<=n-1;i++){
13 int v=z[i];
14 int I=i;
15 for (int j=i+1;<=n;j++)
16 {
17 if(z[j]<v)
18 {
19 v=z[j];
20 I=j
21 {
22 {;
23 if (I==1)
24 continue;
25 else{
26 z[I]=z[i]
27 z[i]=v;
28 }}
29 return 0;
30 }
No
Yes
No
Yes
No
Yes
Development of a hybrid method for calculation of software complexity
Системні дослідження та інформаційні технології, 2022, № 2 41
1 Public Class
2 {
3
4 Public static void main
(String[]args)
5
6 {
7 int arr[10]={1,9,0,5,6,7,
8,2,4,3};
8 int length = 10;
9 int result[]=new int (10);
10 result=QuickSort (arr,0,
length-1);
11 }
12
13 Static int [] QuickSort (int
[]a,int r)
14
15 if (I<r)
16
17 {
18 int i=I;
19 int j=r;
20 int k = (int)((I+r)/2);
21 int pilot = a[k];
22
23 do
24 {
25 while (a[i].less (pilot))
26 i++
27 while (pilot.less (a[i]))
28 j--;
29
30 if (i<=j)
31 {
32 int t = a[i];
33 a[i] = a[j];
34 a[j] = t;
35
36 i++;
37 j++;
38
39 }
40 }
41 while (i<j)
42
43 a=QuickSort{a,i,j};
44 a=QuickSort{a,i,r};
45 }
46
47 return a;
48
49 } //end of QuickSort
50
51
52 } // end of class QS
1 #include<iostream>
2 using namespace std;
3
4 int main()
5
6 {
7
8 int A[]={}1,9,0,5,6,7,8,2,3,4
9 intlength = 10;
10 quickSort(A,0,length-1);
11
12
13 void quicksort(int A[],int
F,int L)
14 {
15 int pivotIndex;
16
17 if (F<L)
18
19 Partition(A,F,L,pivotindex;
20 quicksort(A,F,pivotindex-1);
21 quicksort(A,pivotindex+1,L);
22 }
23 }
24
25 void partition(int A[],int F,
int L,int & pivotindex)
26 {
27 int piv = A[F];
28 int lastS1 = F;
29 int firstUnknown = F+1;
30
31 for(firstUnknown = F+1;++
firstUnknown)
32 {c
33 if (A[firstUnknown]<pivot)
34 {++lastS1;
35 Swap(A[firstUnknown],
A[lastS1]);
36 }
37 }
38 Swap(A[F],A[lastS1]);
39 pivotindex=lastS1;
40
41 }
42
43 void Swap(int & x,int & y)
44 {
45 int temp = x;
46 X=y;
47 Y-temp;
48 }
Fig. 7. Quick sort implementation code in C++ Fig. 8. Quick sort implementation code in Java
T.H. Kazimov, T.A. Bayramova
ISSN 1681–6048 System Research & Information Technologies, 2022, № 2 42
T a b l e 3 . Software complexity of C++ and Java programs
Complexity
metrics
Calculation C++ Java
LOC Number of lines of code 48 52
MacCabe
For a single program
2 neC
e=28, n=26 4 e=29, n=25 6
Halstead )(*)2( 221 hNhD
N1=142
N2=57
h1=25
h2=21
33,92 39,75
Hybrid
n
i
imiE
0
)1(
rnNEC )(loglog 22
N=142,
n=3, r=0,
m1=1 (i=1),
m0=2 (i=0)
9,11
N=133, n=5,
r=0, m0=3
(i=0), m1=1
(i=1), m2=1
(i=2)
10
CONCLUSİON
Software complexity metrics are one of the key aspects of software process
management. Complexity makes software difficult to understand, which creates
problems when maintaining a software system and adding new functions to it. In
the articles of recent years, various methods of solving this problem have been
proposed. Numerous studies in the field of software complexity metrics suggest
that there is no universal metric for assessing the complexity of any program
code. The use of any metric, hybrid metric, or multiple metrics depends on the
specific problem. This article proposes a new hybrid method for calculating the
complexity of software, the effectiveness of which is substantiated by
experiments. The proposed metric makes it possible to clarify (due to additional
indicators) the complexity of software modules of a large class separately from
the known metrics. Based on the above, it is concluded that software metrics are
important at the stages of the project life cycle and, with early application of
program metrics, help to largely overcome the presence of errors and, thus, save
time and money.
REFERENCES
1. T. DeMarco, Controlling Software Projects. Yourdon Press, New York, 1982, 816 p.
2. S. Reddivari and J. Raman, “Software Quality Prediction: An Investigation Based on
Machine Learning”, 2019 IEEE 20th International Conference on Information Reuse
and Integration for Data Science (IRI), pp. 115–122. doi: 10.1109/IRI.2019.00030.
3. A.V. Smirnov, “Methods for assessing and managing the quality of software”, Izves-
tia ETU “LETI” no. 2, pp. 20–25, 2019.
4. I. Somerville, Software engineering, 10th edition. Pearson, 2015, 816 p.
5. S.A. Yaremchuk, “Method for estimating the number of software defects using com-
plexity metrics”, Radioelectronic and computer systems, no. 5, pp. 212–218, 2012.
6. P.A. Laplante, Dictionary of computer science, engineering and technology. CRC
Press, 2017, 560 p.
7. J. Rashid, T. Mahmood, M.W. Nisar, “A Study on Software Metrics and its Impact
on Software Quality”, Technical Journal, University of Engineering and Technology
(UET), Taxila, Pakistan, vol. 24, no. 1, pp. 1–14, 2019.
Development of a hybrid method for calculation of software complexity
Системні дослідження та інформаційні технології, 2022, № 2 43
8. T. Honglei, S. Wei and Z. Yanan, “The Research on Software Metrics and Software
Complexity Metrics”, 2009 International Forum on Computer Science-Technology
and Applications, pp. 131–136, 2009. doi: 10.1109/IFCSTA.2009.39.
9. T.H. Kazimov and T.A. Bayramova, “Evaluating Key Performance Indicators for
Software Development”, IV International Congress on New Trends in Science, En-
gineering and Tehcnology, pp. 99–105, 2020.
10. Shweta, S. Sharma, and R. Singh, “Analysis of correlation between software com-
plexity metrics”, International Journal of Innovative Science, Engineering & Tech-
nology, vol. 2 issue 8, August 2015.
11. S. McIntosh et al., “An empirical study of the impact of modern code review prac-
tices on software quality”, Empirical Software Engineering, 21, pp. 2146–2189,
2016. Available: https://doi.org/10.1007/s10664-015-9381-9.
12. S. Bhatia and J. Malhotra, “A survey on impact of lines of code on software com-
plexity”, 2014 International Conference on Advances in Engineering & Technology
Research (ICAETR – 2014), pp. 1–4, 2014. doi: 10.1109/ICAETR.2014.7012875.
13. A. Ghazarian, “A Theory of Software Complexity”, 2015 IEEE/ACM 4th SEMAT
Workshop on a General Theory of Software Engineering, pp. 29–32, 2015. doi:
10.1109/GTSE.2015.11.
14. M. Khan, F. Ahmad, and M.A. Khanum, “Literature review on software complexity,
software usability and software deliverability”, International Journal of Advanced
Research in Computer Science, vol. 9, no. 2, pp. 438–441, 2018. doi:
10.26483/ijarcs.v9i2.5853.
15. M. Alenezi and M. Zarour, “On the relationship between software complexity and
security”, International Journal of Software Engineering & Applications (IJSEA),
vol. 11, no. 1, pp. 51–60, 2020.
16. S.A. Abdulkareem and A.J. Abboud, “Evaluating Python, C++, JavaScript and Java
Programming Languages Based on Software Complexity Calculator (Halstead Met-
rics)”, 2nd International Scientific Conference of Engineering Sciences (ISCES
2020) 16th-17th December, Diyala, Iraq, IOP Conference Series: Materials Science
and Engineering, vol. 1076, pp. 1–9, 2020. doi: 10.1088/1757-899X/1076/1/012046.
17. Nikhil Govil, “Applying Halstead Software Science on Different Programming Lan-
guages for Analyzing Software Complexity”, Proceedings of the Fourth Interna-
tional Conference on Trends in Electronics and Informatics (ICOEI 2020), pp. 939–
943, 2020. doi: 10.1109/ICOEI48184.2020.9142911
18. Y. Javed, M. Alenezi, M. Akour, and A. Alzyod, “Discovering the relationship be-
tween software complexity and software vulnerabilities”, Journal of Theoretical and
Applied Information Technology, vol. 96, no. 14, pp. 4690–4698, 2018.
19. Y. Shin, “Exploring Complexity Metrics as Indicators of Software Vulnerability”, in
the 3rd International Doctoral Symposium on Empirical Software Engineering,
Kaiserslautern, Germany, October 8, 2008.
20. I. Chowdhury and M. Zulkernine, “Can complexity, coupling, and cohesion metrics
be used as early indicators of vulnerabilities?”, Proceedings of the 2010 ACM Sym-
posium on Applied Computing – SAC ‘10, 2010. doi:10.1145/1774088.1774504.
21. O. Masmali and O. Badreddin, “Towards a Model-based Fuzzy Software Quality
Metrics”, 8th International Conference on Model-Driven Engineering and Software
Development, pp.139–148, 2020.
22. Tong Yi and Chun Fang, “A Novel Method of Complexity Metric for Object-
Oriented Software”, International Journal of Digital Multimedia Broadcast-
ing, vol. 2018, pp. 1–9, 2018. Available: https://doi.org/10.1155/2018/7624768
23. S. Moshin Reza, M. Mahfujur Rahman, H. Parvez, O. Badreddin, and S. Al Mamun,
“Performance Analysis of Machine Learning Approaches in Software Complexity
Prediction”, Proceedings of International Conference on Trends in Computational
and Cognitive Engineering. Advances in Intelligent Systems and Computing, vol.
1309, pp. 27–39, 2021. doi: 10.1007/978-981-33-4673-4_3.
24. P. Morrison, K. Herzig, B. Murphy, and L. Williams, “Challenges with applying
vulnerability prediction models”, in Proceedings of the 2015 Symposium and Boot-
camp on the Science of Security, pp. 1–9, 2015.
T.H. Kazimov, T.A. Bayramova
ISSN 1681–6048 System Research & Information Technologies, 2022, № 2 44
25. Mohd. Kamran Khan et al., “Literature review on software complexity, software
Usability and software deliverability”, International Journal of Advanced Research
in Computer Science, 9 (2), pp. 438–441, 2018.
26. M. Devon Simmonds, “Complexity and the Engineering of Bug-Free Software”,
Proceedings of the International Conference on Frontiers in Education: Computer
Science and Computer Engineering (FECS), Athens, pp. 94–100, 2018.
27. T. Hariprasad, G. Vidhyagaran, K. Seenu, and C.Thirumalai, “Software complexity
analysis using halstead metrics”, 2017 International Conference on Trends in
Electronics and Informatics (ICEI), Tirunelveli, pp. 1109–1113, 2017.
doi: 10.1109/ICOEI.2017.8300883.
28. A. Athar Khan, M. Amjad, Sajeda M. Amralla, and Tahera H. Mirza, “Comparison
of Software Complexity Metrics”, International Journal of Computing and Network
Technology, no. 1, pp.19–26, 2016.
29. G.R. Choudhary, S. Kumar, K. Kumar, A. Mishra, and C. Catal, “Empirical analysis
of change metrics for software fault prediction”, Computers & Electrical Engineer-
ing, vol. 67, pp. 15–24, 2018. Available: https://doi.org/10.1016/ j.compeleceng.
2018.02.043.
30. N. Govil, “Applying Halstead Software Science on Different Programming Lan-
guages for Analyzing Software Complexity”, 2020 4th International Conference on
Trends in Electronics and Informatics (ICOEI)(48184), pp. 939–943, 2020. doi:
10.1109/ICOEI48184.2020.9142911.
31. L. Pudovkina and V. Sinyaiev, “Applying empirical models and halstead metrics to
evaluate the quality of application software”, Open Information and Computer Inte-
grated Technologies, no. 86, pp.190–197, 2019. doi: 10.32620/oikit.2019.86.14
32. T.J. McCabe, “A Complexity Measure”, in IEEE Transactions on Software Engi-
neering, vol. SE-2, no. 4, pp. 308–320, 1976. doi: 10.1109/TSE.1976.233837.
33. H. Liu, X. Gong, L. Liao and B. Li, “Evaluate How Cyclomatic Complexity Changes
in the Context of Software Evolution”, 2018 IEEE 42nd Annual Computer Software
and Applications Conference (COMPSAC), pp. 756–761, 2018. doi:
10.1109/COMPSAC.2018.10332.
Received 22.09.2021
INFORMATION ON THE ARTICLE
Tofig Kazimov (Hasanaga), ORCID: 0000-0001-9245-6731, Institute of Information
Technology of Azerbaijan National Academy of Sciences, Azerbaijan, e-mail: to-
fig@mail.ru
Tamilla Bayramova (Adil), ORCID: 0000-0002-8377-3572, Institute of Information
Technology of Azerbaijan National Academy of Sciences, Azerbaijan, e-mail:
toma_b66@mail.ru
РОЗРОБЛЕННЯ ГІБРИДНОГО МЕТОДУ ОБЧИСЛЕННЯ СКЛАДНОСТІ
ПРОГРАМНОГО ЗАБЕЗПЕЧЕННЯ / Т.Г. Казімов, Т.А. Байрамова
Анотація. Застосування показників програмного коду дозволяє розробникам
програмного забезпечення і керівникам проектів оцінювати різні функції про-
грамного забезпечення, яке буде створено або вже існує, прогнозувати робоче
навантаження, визначати складність і надійність програмного забезпечення і
надавати кількісні характеристики якості розроблюваних програмних систем.
У працях останніх років пропонуються різні методи вирішення цієї проблеми.
Утім досі немає ефективного підходу до вимірювання складності програмного
забезпечення. Подано стислий огляд існуючих метрик складності програмного
забезпечення. Запропоновано новий гібридний метод обчислення складності
програмного забезпечення, який об’єднує ключові характеристики метрик Хо-
лстеда, Маккабі і SLOC, а також дозволяє більш ефективно оцінювати склад-
ність.
Ключові слова: програмна інженерія, складність програмного забезпечення,
метрики складності, гібридний метод.
|
| id | journaliasakpiua-article-240177 |
| institution | System research and information technologies |
| keywords_txt_mv | keywords |
| language | English |
| last_indexed | 2025-07-17T10:27:30Z |
| publishDate | 2022 |
| publisher | The National Technical University of Ukraine "Igor Sikorsky Kyiv Polytechnic Institute" |
| record_format | ojs |
| resource_txt_mv | journaliasakpiua/f8/49adc79aaa7677b6054c1e211f4204f8.pdf |
| spelling | journaliasakpiua-article-2401772022-10-17T22:12:39Z Development of a hybrid method for calculation of software complexity ИССЛЕДОВАНИЕ МЕТРИК СЛОЖНОСТИ ПРОГРАММНОГО ОБЕСПЕЧЕНИЯ Розроблення гібридного методу обчислення складності програмного забезпечення Kazimov, Tofig Bayramova, Tamilla software engineering software complexity complexity metrics hybrid method програмна інженерія складність програмного забезпечення метрики складності гібридний метод The use of code metrics allows software developers and project managers to evaluate various features of the software (to be built or already in existence), predict workload, determine software complexity and reliability, and quantify the quality of software systems being developed. Articles written in recent years have proposed various methods for solving this problem. However, there is still no very effective approach to measuring software complexity. This article provides a brief overview of existing software complexity metrics and proposes a new hybrid method for computing software complexity. The proposed hybrid method for evaluating software complexity combines the key features of the Halsted, Maccabe, and SLOC metrics and also allows for a more efficient assessment of complexity. Применение показателей программного кода позволяет разработчикам программного обеспечения и руководителям проектов оценивать различные функции программного обеспечения, которое будет создано или уже существует, прогнозировать рабочую нагрузку, определять сложность и надежность программного обеспечения и предоставлять количественные характеристики качества разрабатываемых программных систем. В статьях, написанных в последние годы, предлагаются различные методы решения этой проблемы. Однако до сих пор нет очень эффективного подхода к измерению сложности программного обеспечения. В этой статье дается краткий обзор существующих метрик сложности программного обеспечения и предлагается новый гибридный метод для вычисления сложности программного обеспечения. Предлагаемый гибридный метод оценки сложности программного обеспечения объединяет ключевые характеристики метрик Холстеда, Маккабе и SLOC, а также позволяет более эффективно оценивать сложность. Застосування показників програмного коду дозволяє розробникам програмного забезпечення і керівникам проектів оцінювати різні функції програмного забезпечення, яке буде створено або вже існує, прогнозувати робоче навантаження, визначати складність і надійність програмного забезпечення і надавати кількісні характеристики якості розроблюваних програмних систем. У працях останніх років пропонуються різні методи вирішення цієї проблеми. Утім досі немає ефективного підходу до вимірювання складності програмного забезпечення. Подано стислий огляд існуючих метрик складності програмного забезпечення. Запропоновано новий гібридний метод обчислення складності програмного забезпечення, який об’єднує ключові характеристики метрик Холстеда, Маккабі і SLOC, а також дозволяє більш ефективно оцінювати складність. The National Technical University of Ukraine "Igor Sikorsky Kyiv Polytechnic Institute" 2022-08-30 Article Article application/pdf https://journal.iasa.kpi.ua/article/view/240177 10.20535/SRIT.2308-8893.2022.2.02 System research and information technologies; No. 2 (2022); 32-44 Системные исследования и информационные технологии; № 2 (2022); 32-44 Системні дослідження та інформаційні технології; № 2 (2022); 32-44 2308-8893 1681-6048 en https://journal.iasa.kpi.ua/article/view/240177/261587 |
| spellingShingle | програмна інженерія складність програмного забезпечення метрики складності гібридний метод Kazimov, Tofig Bayramova, Tamilla Розроблення гібридного методу обчислення складності програмного забезпечення |
| title | Розроблення гібридного методу обчислення складності програмного забезпечення |
| title_alt | Development of a hybrid method for calculation of software complexity ИССЛЕДОВАНИЕ МЕТРИК СЛОЖНОСТИ ПРОГРАММНОГО ОБЕСПЕЧЕНИЯ |
| title_full | Розроблення гібридного методу обчислення складності програмного забезпечення |
| title_fullStr | Розроблення гібридного методу обчислення складності програмного забезпечення |
| title_full_unstemmed | Розроблення гібридного методу обчислення складності програмного забезпечення |
| title_short | Розроблення гібридного методу обчислення складності програмного забезпечення |
| title_sort | розроблення гібридного методу обчислення складності програмного забезпечення |
| topic | програмна інженерія складність програмного забезпечення метрики складності гібридний метод |
| topic_facet | software engineering software complexity complexity metrics hybrid method програмна інженерія складність програмного забезпечення метрики складності гібридний метод |
| url | https://journal.iasa.kpi.ua/article/view/240177 |
| work_keys_str_mv | AT kazimovtofig developmentofahybridmethodforcalculationofsoftwarecomplexity AT bayramovatamilla developmentofahybridmethodforcalculationofsoftwarecomplexity AT kazimovtofig issledovaniemetriksložnostiprogrammnogoobespečeniâ AT bayramovatamilla issledovaniemetriksložnostiprogrammnogoobespečeniâ AT kazimovtofig rozroblennâgíbridnogometoduobčislennâskladnostíprogramnogozabezpečennâ AT bayramovatamilla rozroblennâgíbridnogometoduobčislennâskladnostíprogramnogozabezpečennâ |