Adaptive Test Suits Generation for Self-Adaptive Systems Using SPEA2 Algorithm
, Mohamed K. Nour 1, Saud S. Alotaibi 2, Mohammad Jabed Hussain 2, Syed Mutiullah Hussaini 1 and Atif Naseer 3
Round 1
Reviewer 1 Report
The review of the manuscript: applsci-2572150
Full Title: Adaptive Test Suits Generation for Large Configurable Systems using
SPEA2 Algorithm
General remarks
The topic of the manuscript is attractive and might be of interest for the readers of Applied Sciences. Authors introduced an evolutionary framework that supports adaptive testing for Self-18 adaptive systems. They proposed approach using the SPEA2 algorithm facilitates both the execution and adaptation of run-time testing operations.
Remarks by sections
· The Introduction section is well presented and organized. There are cited all necessary previously published papers.
· The Background section contains all relevant previous work.
· The Proposed Approach-SPEA2-SAS section is well explained.
· The Experiments and Results section is clear, with good explained Tables and Figures.
· The Discussion and Conclusion sections are concise.
The issues:
· The abbreviation of SPEA2 is not presented in the text.
· Page 5, Procedure SPEA2, first line, what the triangle stands for?
Recommendation: manuscript is acceptable for publication after these minor changes.
Author Response
The article has been updated as per Reviewer's instructions.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
the paper is promising, however, the following comments should be considered:
1. the paper was based on some challenges mentioned in the introduction. the authors mentioned references to confirm the existence of these challenges. however, these references are old (2013, 2011,..) which leads to the need for more reading by the authors in recent research to confirm that these challenges are still not tackled.
2. in general, the references in the research need to be more recent
3. the research mentions that selected features can change at runtime. there is no explanation of how this change will occur (especially, since the features selection step is part of the static phase). this is also conflicting with the part "SAS is frequently built to search for the feature combinations dynamically"
4. I have to mention again that even the proposed adaptation is based on old references (2013 and older) which is not acceptable.
5. optimization is based on SPEA2. have the authors investigated the recent enhancements of the algorithm?
6. although figure 3 is generalized, in the text, it is mentioned that Figure 3 shows how SPEA2 contributes to the optimization task for both cost and time features. (more explanation is required)
7. the experiment section needs to include different experiments with manipulating the evaluation parameters. results should be illustrated for more clarification. a comparison before and after the model contribution should be discussed to highlight the paper's contribution.
Author Response
The article has been updated as per Reviewer's instructions.
Author Response File: Author Response.pdf
Reviewer 3 Report
1. The structure of this article is not very reasonable. The first two parts should not be introduction and background. The introduction section introduces the background and innovation points, and then the related works section introduces related work.
2. The experimental part does not define relevant indicators to conduct quantitative experiments, and there are no relevant comparative experiments.
state-of-the-art in line 100
Author Response
The article has been updated as per Reviewer's instructions.
Author Response File: Author Response.pdf
Reviewer 4 Report
The paper presents a framework called SPEA2-SAS that uses a multi-objective evolutionary algorithm (MOEA) to optimize self-adaptive software (SAS) at runtime. The framework automatically converts the SAS feature model into the MOEA context and takes into account the dependencies and trade-offs among the features. The framework also uses an elitist chromosome representation and dependency-aware operators to guide the search for better solutions. The article evaluates the framework on two real-world SAS with different parameters and compares it with state-of-the-art approaches. The results show that SPEA2-SAS can produce better and more balanced solutions with reasonable overhead.
Comments:
Chapter 5.1. State of Art Comparison shouldn't be in the discussion chapter. This comparison should be moved to the Introduction or Background sections!
Figures 1, 3, and 4 are not properly cited. Are these figures the work of the authors?
References 27 and 35 are highlighted in red. Why?
Author Response
The article has been updated as per Reviewer's instructions.
Author Response File: Author Response.pdf
Round 2
Reviewer 2 Report
I will build this review according to the highlighted parts in the revised version. Although all the responses are supposed to be considered, however, I can not find this consideration in the revised version.
1. the authors mention that recent references have been added. However, they are a maximum of 2021. I have to ask if there has been no work in this direction for the past two years?
2. the response to point 3 should be included in the paper.
3. although it is mentioned in point 6 that there is an additional explanation, I could not follow where this explanation exists in the paper.
4. point 7 mentions an added contribution in the conclusion, while the only added part is an additional future work. By the contribution, I do not mean the authors' effort in the paper; I mean the contribution of the research in the field.
Author Response
The article has been updated according to the reviewer's comments.
Author Response File: Author Response.pdf
Reviewer 3 Report
accept
accept
Author Response
The article has been updated according to the reviewer's comments.
Author Response File: Author Response.pdf
Round 3
Reviewer 2 Report
the authors covered the comments
Author Response
All the changes are made as per reviewers' instructions.
Author Response File: Author Response.pdf
