AN IMPROVED REPOSITORY STRUCTURE TO IDENTIFY, SELECT AND INTEGRATE COMPONENTS IN COMPONENT-BASED DEVELOPMENT

International Journal of Software Engineering & Applications (IJSEA), Vol.13, No.6, November 2022
DOI: 10.5121/ijsea.2022.13603 31
AN IMPROVED REPOSITORY STRUCTURE TO
IDENTIFY, SELECT AND INTEGRATE COMPONENTS
IN COMPONENT-BASED DEVELOPMENT
Muhammad Khamis Dauda, Reda M Salama and Rizwan Qureshi
Faculty of Computing and Information Technology,
King Abdulaziz University, Jeddah. Saudi Arabia
ABSTRACT
An ultimate goal of software development is to build high quality products. The customers of software
industry always demand for high-quality products quickly and cost effectively. The component-based
development (CBD) is the most suitable methodology for the software companies to meet the demands of
target market. To opt CBD, the software development teams have to customize generic components that are
available in the market and it is very difficult for the development teams to choose the suitable components
from the millions of third party and commercial off the shelf (COTS) components. On the other hand, the
development of in-house repository is tedious and time consuming. In this paper, we propose an easy and
understandable repository structure to provide helpful information about stored components like how to
identify, select, retrieve and integrate components. The proposed repository will also provide previous
assessments of developers and end-users about the selected component. The proposed repository will help
the software companies by reducing the customization effort, improving the quality of developed software
and preventing integrating unfamiliar components.
KEYWORDS
CBD, COTS, Third Party Components, Repository, Quality Attributes.
1. INTRODUCTION
Every software organization aims to produce or to acquire qualitative software, to save time, cost
and meet the market demand. During the analysis phase, suitable components are searched from
both the in house repository and third party repository. A lot of time is spent in identification,
selection, and analysis, which increases the customization and integration cost because the
component’s structure did not provide any insight to help the team with some basic information
including quality constituents or attribute and the development team has no direction [1]. A lot of
time is also spent during customization and integration of reusable components. It is highly likely
that more suitable components might not be identified resulting in wastage of time, cost, effort and
resources. There is a need to provide a structure to evaluate quality attributes of commercial off
the shelf (COTS) or in house components during the selection phase. There is also a need to
monitor the version control system, end-user feedback, quality of service (QoS) issues and
maturity of the component vendor [2].
The paper is further arranged as follows. Section II covers the related work. The problem
statement is described in section III. Section IV illustrates the proposed solution. The proposed
solution is validated in section V. The recommendations and future work are covered in section
VI.

32
2. RELATED WORK
The idea of component-based software engineering (CBSE) among the software community is
not new and the software development companies are practicing it from more than two decades.
Quality evaluation and metrics are required to know for every individual component before its
reuse. However, a formal direct and an indirect component coupling metrics are proposed to
measure the quality of components with respect to complexity and performance [3]. The research
investigates the risk-management activities and their correlations with the occurrences of typical
risks in the development of COTS systems [4]. It is achieved by exploring the occurrences of
typical risks in COTS systems and furtherly the effectiveness of the risk reduction activities are
compared.
The aim of the proposed research is to improve security and quality concerns of open-source
software systems [5]. The major challenges of open source development are addressed. A model
is proposed to assure quality [5]. The study explores the current state of tool building in the
reverse engineering domain intending to improve upon the practice to a predictable format [6].
The symptoms of code smells are poor design and implementation choices [7]. The high cost of
maintainability and customization is due to smelly code. A model is proposed to identify poor
quality software, bug prediction, and bug classification using F-measures. It is important to
understand the implication of choosing a suitable component from a third-party repository
because of the trade-off in quality [8]. Software Engineering taught programs do not teach how to
ensure that COTS components are not compromised from production to integration [9]. Data are
generated during component execution, which can be distilled and mined [10]. The study in
explores the challenges of DevOps including performance and quality [11]. A quality assurance
model is proposed to implement during the phases of analysis, development, certification,
customization, design, integration, testing, and maintenance [11].
Table 1. Limitations of Related Work.
The study in [12] focuses on software refactoring and quality enhancement. A pilot survey on
data analysis claimed that the construction of a new framework for healthcare COTS evaluation
and selection is necessary [13]. Most scientific software offer performance and maintainability
Paper Title Limitation
A Review of Component Coupling Metrics for
Component-Based Development [3].
The idea is proposed to establish to
measure the effectiveness of reusable
components.
A State-of-the-Practice Survey of Risk
Management in Development with Off-the-Shelf
Software Components [4].
The results show that there are still
unexplained risk-related factors in the
proposed risk-reduction activities.
A Model for Quality Assurance of OSS
Architecture [5].
The quality assurance model is very hard
to use in software engineering.
Building Reverse Engineering Tools with
Software Components: Ten Lessons Learned [6].
More case studies are needed to
conclude the results
Toward a Smell-Aware Bug Prediction Model
[7].
Only Apache data sets are used to
validate the model.
Choosing Component Origins for Software
Intensive Systems: In-House, COTS, OSS, or
Outsourcing? —A Case Survey [8].
There is a need to conduct empirical
studies to generalize the results.

33
quality attributes at the highest priority [14]. As soon as new requirements are required to
implement in a system, the developers must identify their attributes and impact on the existing
components of respective system [15]. The study uses 18 matrices and 6 NLP techniques to
measure and identify the semantic similarities of a text. The experimental results show that the
accuracy of predicting impacted classes are increased more than 60 percent. High software
quality is mandatory in an organization to avoid costly patching [16]. There is a need of a
standardized quality check to maintain consistency and reliability among recent technologies
[17]. It is a common dilemma that increasing demand of working software and integration factors
led to forget about the quality of selected components. Table 1 illustrates the limitations of
related work [3-8].
3. PROBLEM DEFINITION
According to the proposed extended CBD in [18], ‘analysis, selection, and risk management’
phase collects detailed specifications of the system to be developed.
Software industries face difficulties in selecting suitable components that will match the
requirements of the new system. Many research studies suggest the need to find a suitable way
for component selection.
How to propose a repository structure to facilitate easy identification, selection and integration of
a reusable component?
4. THE PROPOSED SOLUTION
We propose a structure for software industries regardless of the methodology that will contribute
to solving the problem of identification, selection, and acquiring a qualitative component from
repositories as shown in figure 1.
A component possesses specific quality attribute like its domain, type of design pattern, and the
intended users. To make our structure flexible and easily adaptable to organizations, we limited
our structure to the following basic information that provides the ultimate goal needed.
Figure 1. The Proposed Structure showing Quality attributes with team rating

34
Quality attributes of a component will serve two purposes:
 the maturity level;
 easy identification and retrieval from the repository.
A relative number will be assigned to each quality attribute by the development team during
meetings from one to ten. Customer collaboration and feedback will help to rate a component. As
a result, the proposed structure fulfills two functions related to quality:
 quality views based on the user;
 quality views based on the team.
A component always belongs to a specific design pattern and it is stored in the repository as per
its type. It is extremely helpful for the development team to know that the selected component
belongs to creational, structural or behavioral patterns. It will help to determine the architecture
and degree of customization and integration risks. Estimate the domain of a component, the
development team can reduce customization cost and time. A component can have different
versions that can have different scores and ratings. The trade-off of the quality attribute will be
visible by version control mechanism. It will help to determine, select and acquire the most
feasible component and save time, cost and effort of a team as proposed in figure 2. According to
figure 2, the proposed repository structure will save time, effort and integration cost during the
component acquisition process.
Figure 2. The Proposed Repository Structure to Perform Component Acquisition
5. VALIDATION
There are many techniques available to validate the proposed solution. We found that
questionnaire technique is more suitable to validate our research proposal as per the time
constraints. This paper is an outcome of one of master courses taught at King Abdul-Aziz
University, KSA. The duration of the course is six months and one of the benefits of
questionnaire technique is that it can examine the responses of a relatively large number of
participants in less time. To validate our proposal, we targeted only those participants who are

35
practicing CBD through emails using embedded google sheets. The questionnaire is composed of
twenty-one questions to validate the two goals of proposed solution.
Table 2. Likert scale used to evaluate the questionnaire
Strongly Agreed 5
Agreed 4
Neither Agreed Nor Disagreed 3
Disagreed 2
Strongly Disagreed 1
Goal 1 is to identify the organizational acceptability of the proposed repository structure and it is
evaluated against eleven questions. The result of the acceptability will indicate the positive
impact and usefulness of the structure because organizations (stakeholders, investors) cannot
accept any structure without vividly seeing its positive results with respect to identify, select and
acquire a component.
Goal 2 is to inquire that how much proposed repository structure facilitates to the development
team and it is assessed using ten questions. Goal 2 will help the researchers to analyze the
effectiveness of the proposed structure to adapt and integrate a reusable component.
The questions are evaluated using likert scale as shown in Table 2.
5.1. Goal 1: Determine the Organization Acceptability of the Proposed Structure
The questions in this goal represent the acceptability of the proposed repository structure to
identify, select and retrieve the reusable components. Table 3 shows that 34% are strongly agreed
whereas 28% are agreed. Furthermore, 21% are neither agreed nor disagreed. However, 12% are
disagreed while other 5% are strongly disagreed. Figure 3 depicts the same result graphically.
Table 3. Analysis Result of Goal 1
Q. No. Strongly
Disagreed
Disagreed Neutral Agreed Strongly
Agreed
Q1. 0 1 3 7 11
Q2. 0 1 2 5 14
Q3. 2 7 7 3 3
Q4. 1 10 6 1 4
Q5 6 6 4 4 2
Q6 2 7 5 5 3
Q7 1 3 8 5 5
Q8 1 1 4 8 8
Q9 0 1 3 8 10
Q10 0 2 3 12 5
Q11 0 1 3 6 12
Total 13 30 48 64 77
Avg. in % 5 12 21 28 34

36
Figure 3. Cumulative Frequency Results of Goal 1
5.2. Goal 2: Adaption of Proposed Structure by Development Team
According to Table 4, the analysis of the results show that 35% are agreed and 31% are strongly
agreed that the proposed structure facilitates to development teams to adapt and integrate the
reusable components. Whereas 13% are disagreed and 1% are strongly disagree with the goal 2.
Also, 20% are remained neutral. The results of Table 4 are presented graphically in figure 4.
Figure 4. Cumulative Frequency Results of Goal 2
Table 4. Analysis result of goal 2
Q. No. Strongly
Disagreed
Agreed
Q1. 0 3 5 4 10
Q2. 0 3 4 9 6
Q3. 1 3 7 7 4
Q4. 1 4 8 4 6
Q5 1 4 3 8 6
Q6 0 2 7 9 4
Q7 1 1 1 10 9
Q8 0 3 3 7 9
Q9 0 3 4 9 6
Q10 0 4 6 7 5
Total 4 30 44 74 65
Avg. in % 1 13 20 35 31
5
12
21
28
34
0
10
20
30
40
1 2 3 4 5
Frequency
in
percetage
likert Scale
1
13
20
35
31
0
10
20
30
40
1 2 3 4 5
Frequency
in
percentage
likert Scale

37
5.3. Cumulative Analysis of Two Goals
Table 5 shows the cumulative analysis of two goals to evaluate the proposed research.
Table 5. Cumulative Analysis of Two Goals
Goal No. Strongly
Disagreed
Agreed
Goal 1 5 12 21 28 34
Goal 2 1 13 20 35 31
Total 6 25 41 63 65
Avg. in % 3 12 20 32 33
Figure 5. Cumulative Frequency Results of Two Goals
Figure 5 demonstrates the outcomes of cumulative statistical analysis of two goals. Thirty-two
percentage of the participants are agreed with the proposed repository structure will help the
software development teams during the selection and retrieval of suitable reusable components.
Thirty-three percentage of the respondents are strongly agreed and twenty percentage of the
respondents are neither agreed nor disagreed. Twelve percentage of the respondents are disagreed
and only three percentages are strongly disagreed as shown in Table 5.
6. RECOMMENDATION AND FUTURE WORK
There are many version control libraries but those repositories do not cover all the necessary or
required information to identify, select and integrate components using CBD. An improved
repository structure is proposed to solve the problem in hand. BitBucket concentrates more on
integrating and coordinating team while GitHub focuses more on component maturity and
performance. GitLab provides hybrid features by concentering more on managing, organizing
team, security and compliance but still it lacks in primary features mentioned in our proposed
repository structure. The future work is to propose an extension of improve repository to evaluate
the maturity level of each stored component. It will mitigate high cost of maintainability and
decrease time to identify, select and retrieve a component.
7. CONCLUSION
To develop a high quality software, there is a need to accomplish a successful project coping the
requirements of a customer in cost effective way and meeting the timelines. The problem is that
customers are always pushing the software development companies to deliver products fast to take
the competitive benefits. On the other hand, the software development companies want to reuse
3
12
20
32 33
1 2 3 4 5

38
the previously development components to avoid the scratch based development to meetneeds of
demanding market. There is a need of software companies to store the reusable components in a
repository. The repository contains several versions of each reusable component and it is
extremely difficult for the software development team to select the most suitable version of a
component for the project in hand to customize it. This research proposes a repository structure to
facilitate the software development teams to choose the most appropriate component as per the
requirements of a customer. The proposed repository structure will provide ample information
about each version of reusable components so that the software development teams can easily
analyze, select and retrieve reusable components. The core focus of this research is to propose
such a repository structure that will save time, cost, efforts and resources of the software
development companies by easing the components’ identification, selection and retrieval
processes. A questionnaire is used to validate the proposed repository structure and overall sixty-
five percentage of the respondents are agreed with the effectiveness of the proposal.
REFERENCES
[1] Sommerville I., (2016) Software Engineering, Pearson Publisher.
[2] Maximus B. and Pressman R. S., (2019) Software Engineering, McGraw Hill Publisher.
[3] Chen, J., Yeap, W. K. and Bruda S. D. (2009) “A Review of Component Coupling Metrics for
Component Based Development,” WRI World Congress on Software Engineering, pp65-69.
[4] Li, J., Conradi, R., Slyngstad, O. P., Torchiano, M., Morisio, M. and Bunse, C. (2008) “A State-of-the-
Practice Survey of Risk Management in Development with Off-the-Shelf Software Components,”
IEEE Trans. Softw. Eng., Vol. 34, pp271-286.
[5] Kumar, R. and Singh, H. (2012) “A model for quality assurance of OSS architecture,” Sixth Int. Conf.
Software Engineering (CONSEG), pp1-6.
[6] Kienle, H. M. (2007) “Building Reverse Engineering Tools with Software Components: Ten Lessons
Learned,” 14th Working Conference on Reverse Engineering (WCRE 2007), pp289-292.
[7] Palomba, F., Zanoni, M., Fontana, F. A., De Lucia, A. and Oliveto, R. (2019) “Toward a Smell-Aware
Bug Prediction Model,” IEEE Trans. Softw. Eng., Vol. 45, pp194-218.
[8] Petersen, K. et al. (2018) “Choosing Component Origins for Software Intensive Systems: In-House,
COTS, OSS or Outsourcing?—A Case Survey,” IEEE Trans. Soft. Eng., Vol. 44, pp237-261.
[9] Mead, N. R., Kohnke, A. and Shoemaker, D. (2020) “Secure Sourcing of COTS Products: A Critical
Missing Element in Software Engineering Education,” 32nd Conf. Software Engineering Education
and Training (CSEE T), pp1-5.
[10] Liu, C. (2020) “Discovery and Quality Evaluation of Software Component Behavioral Models,” IEEE
Trans. Autom. Sci. Eng., Vol. 18, pp1538-1549.
[11] Mishra, A. and Otaiwi, Z. (2020) “DevOps and software quality: A systematic mapping,” Comput. Sci.
Rev., Vol. 38, pp100308.
[12] Al Dalla,l J. and Abdin, A. (2018) “Empirical Evaluation of the Impact of Object-Oriented Code
Refactoring on Quality Attributes: A Systematic Literature Review,” IEEE Trans. Soft. Eng., Vol. 44,
pp44-69.
[13] Al-Tarawneh, F. H. and Althunibat, A. (2019) “Pilot Study of Healthcare COTS Software Evaluation
and Selection,” Int. Joint Conf. Electrical Engineering and Information Technology (JEEIT), pp311-
314.
[14] Arvanitou, A., Ampatzoglou, A. C. and Carver, J. C. (2021) “Software engineering practices for
scientific software development: A systematic mapping study,” J. Syst. Soft., Vol. 172, pp110848.
[15] Falessi, D., Roll, J., Guo, J. L. C., and Cleland-Huang, J. (2020) “Leveraging Historical Associations
between Requirements and Source Code to Identify Impacted Classes,” IEEE Trans. Soft. Eng., Vol.
46, pp420-441.
[16] Poth, A., Meyer, B., Schlicht, P. and Riel. A. (2020) “Quality Assurance for Machine Learning – an
approach to function and system safeguarding,” 20th Int. Conf. Software Quality, Reliability and
Security (QRS), pp22-29.
[17] Chakraborty, A., Bagavathi, R. and Tomer, U. (2020) “A Comprehensive Decomposition towards the
Facets of Quality in IoT,” Int. Conf. Smart Electronics and Communication (ICOSEC), pp759-764.

39
[18] Qureshi, M. R. J., Barnawi, A., and Talhi, A. A. (2013) “Component based development model for
new and customized software products,” 3rd World Conf. Information Technology,pp1384-1389.
AUTHORS
Muhammad Khamis Dauda is a certified Scrum Master and M.Sc. student at Faculty
of Computing and Information Technology, King Abdul-Aziz University, Saudi
Arabia.
Dr. Reda M Salama is working as an associate professor in King Abdul-Aziz
University, Jeddah, Saudi Arabia.
Dr. Rizwan Qureshi is currently working as a full professor in the Department of IT,
King Abdul-Aziz University, Jeddah, Saudi Arabia. This author is the best researcher
awardees from the Department of Information Technology, King Abdul-Aziz
University in 2013 and 2016.

AN IMPROVED REPOSITORY STRUCTURE TO IDENTIFY, SELECT AND INTEGRATE COMPONENTS IN COMPONENT-BASED DEVELOPMENT

More Related Content

Similar to AN IMPROVED REPOSITORY STRUCTURE TO IDENTIFY, SELECT AND INTEGRATE COMPONENTS IN COMPONENT-BASED DEVELOPMENT (20)

Recently uploaded (20)

AN IMPROVED REPOSITORY STRUCTURE TO IDENTIFY, SELECT AND INTEGRATE COMPONENTS IN COMPONENT-BASED DEVELOPMENT