A Survey on Particle Swarm Optimization for Association Rule Mining

Round 1

Reviewer 1 Report

-The authors studied the utilizing the Particle Swarm Optimization (PSO) technique on the Association Rule Mining problem. They reviewed the existing research results, heuristic algorithms for ARM were classified into three main groups, including biologically inspired, physically inspired, and based on other algorithms.

-The following major corrections are required:

-The introduction section is too general, and it introduces concepts that are well known to the utilizing the Particle Swarm Optimization (PSO) technique on the Association Rule Mining problem. The introduction does not stimulate to go ahead with the remaining of the paper because it does not introduce any really new topic/solution. Furthermore, “the motivation and research challenges” at the introduction section is missing. Please rewrite this section.

-Please provide a section named “Research methodology” as a guideline for searching relevant articles in association rule mining based on PSO including research questions, search keywords, article selection,…

-This survey paper is mainly based on the classification of existing Particle Swarm Optimization (PSO) technique on the Association Rule Mining. However, no motivation and/or justification are provided that explains why such a categorization is important and how this will help/facilitate the researchers of the field.

-Your study is very similar to the following papers. What is the difference between your study and mentioned papers?

https://www.sciencedirect.com/science/article/abs/pii/S1568494609002361

https://link.springer.com/chapter/10.1007/978-3-642-35314-7_46

Lastly, some key papers in the research area are left out and question and require improving the search methodology. Some of the related papers that should have be included are:

https://www.sciencedirect.com/science/article/abs/pii/S092041052200362X

https://www.sciencedirect.com/science/article/abs/pii/S2214212622000989

https://link.springer.com/article/10.1007/s11227-020-03296-w

https://link.springer.com/article/10.1007/s11277-022-09576-3

https://link.springer.com/article/10.1007/s00500-020-05409-2

https://onlinelibrary.wiley.com/doi/abs/10.1002/spe.2641

https://link.springer.com/article/10.1007/s11831-021-09694-4

-Paper needs some revision in English. The overall paper should be carefully revised with focus on the language: especially grammar and punctuation.

-Overall, there are still some major parts that the authors did not explain clearly. As a result, I am going to suggest Major revision the paper in its present form.

Author Response

Response to Reviewer 1 Comments:

Point 1: The introduction section is too general, and it introduces concepts that are well known to the utilizing the Particle Swarm Optimization (PSO) technique on the Association Rule Mining problem. The introduction does not stimulate to go ahead with the remaining of the paper because it does not introduce any really new topic/solution. Furthermore, “the motivation and research challenges” at the introduction section is missing. Please rewrite this section.

Response 1: Thanks for your suggestions. I reworked the introduction, added some key papers in the field, elaborated on and improved the "Motivation and Research Challenges" section in the introduction, which can be found in lines 62–66 of the article. In addition, we added the research methodology, including the choice of papers, based on the reviewers' comments, to make the introduction more complete and to guide the whole text.

Point 2: Please provide a section named “Research methodology” as a guideline for searching relevant articles in association rule mining based on PSO including research questions, search keywords, article selection,

Response 2: Thanks for your suggestions. I added the section on "Research methodology" to the introduction, including the research questions for writing, which can be found in lines 64-66 of the article. In addition, we used PSO or ARM as keywords to search the latest papers in recent years, removed the papers that did not meet the requirements under the condition of low citation and weak relevance, and used the excellent remaining papers as references for the article's exposition, which can be found in lines 67–69 of the article.

Point 3: This survey paper is mainly based on the classification of existing Particle Swarm Optimization (PSO) technique on the Association Rule Mining. However, no motivation and/or justification are provided that explains why such a categorization is important and how this will help/facilitate the researchers of the field.

Response 3: Thanks for your suggestions. Owing to recent papers reviewed only a small fraction of PSO-based ARM algorithms, but there have been many papers published in recent years. Therefore, we aimed to provide a more comprehensive and multi-faceted survey of emerging research. In this paper, we divide the ARM methods based on heuristic algorithms into three main groups: bio-inspired, physics-inspired, and other algorithms, and provide a paragraph describing the algorithms and a summary of their features in each group. We expect this survey can provide a reference for researchers in the field to help them understand the state-of-the-art PSO-based ARM algorithms.

Point 4: Your study is very similar to the following papers. What is the difference between your study and mentioned papers?

https://www.sciencedirect.com/science/article/abs/pii/S1568494609002361

https://link.springer.com/chapter/10.1007/978-3-642-35314-7_46

Response 4: Thanks for your suggestions. In comparison to the preceding papers, this paper focuses on the research status of ARM from the standpoint of the PSO algorithm and summarizes and composes the ARM from three perspectives: mode types, evaluation metrics, and comparison with other algorithms. In addition, this paper also describes various application areas such as privacy protection and recommender systems to provide a reference for researchers in the field to help them understand the state-of-the-art PSO-based ARM algorithms.

Point 5: some key papers in the research area are left out and question and require improving the search methodology. Some of the related papers that should have be included are:

https://www.sciencedirect.com/science/article/abs/pii/S092041052200362X

https://www.sciencedirect.com/science/article/abs/pii/S2214212622000989

https://link.springer.com/article/10.1007/s11227-020-03296-w

https://link.springer.com/article/10.1007/s11277-022-09576-3

https://link.springer.com/article/10.1007/s00500-020-05409-2

https://onlinelibrary.wiley.com/doi/abs/10.1002/spe.2641

https://link.springer.com/article/10.1007/s11831-021-09694-4

Response 5: Thanks for your suggestions. I have added the relevant papers that need to be added to the article based on your suggestions in the introduction, Section 2.1.1, Section 4.2.1, and Section 4.2.3 of the article. Thank you again. Because of your valuable comments, the literature section of the article seems more complete.

Point 6: Paper needs some revision in English. The overall paper should be carefully revised with focus on the language: especially grammar and punctuation.

Response 6: Thanks for your suggestions. The English expressions and grammatical usage of this paper were revised by checking the whole text. We will also carefully check the grammar of English in our future work, and your advice will be an important guide for my thesis writing skills and scientific work.

Author Response File: Author Response.pdf

Reviewer 2 Report

The topic is interesting to the relevant research field, however, the novel feature of the work is not clear showed.  From this perspective, it needs minor revision. The authors have conducted a brief work but some important things should be included in the revised manuscript. Only then the work can be accepted

1) Abstract should include research purposes. The abstract is too general, the abstract should contain significant results with a certain value or percentage. It needs need to be revised and rewritten.

2) In general the section of Introduction is well organized, but lacks sufficient literature review. The introduction part should be improved further to review the relevant research and highlights the advantages of your work.  

3)  Please read the paper carefully for English language style and accuracy, and make appropriate corrections and changes. With assist of professional experts rectify the grammatical and vocabulary error in the paper wherever necessary.

4) The manuscript has sufficient results, but the final findings should necessarily be validated and compared with the results of standard diesel fuelled engine. Further, the study should clearly indicate the improvement or decline in results. At the end of each section in the results, authors should present concluding remarks with critical discussion in their own words.

5 The real applications and meaning of this study should be further elaborated in the section of Conclusion. It should only contain the significant outcomes and the take home message and novel findings of this manuscript. Authors are advised to rewrite the conclusion section and give some insights about the variations in engine output characteristics numerically.

Author Response

Response to Reviewer 2 Comments:

Point 1: Abstract should include research purposes. The abstract is too general, the abstract should contain significant results with a certain value or percentage. It needs need to be revised and rewritten.

Response 1: Thanks for your suggestions. We strongly agree with your comments and have added the purpose of the study in the abstract section, which can be found in lines 62-66 of the article. However, as our article is a review paper, we are thus unable to add numerical results. Therefore, we hope to avoid this problem in future work to ensure a more convincing article.

Point 2: In general the section of Introduction is well organized, but lacks sufficient literature review. The introduction part should be improved further to review the relevant research and highlights the advantages of your work.

Response 2: Thanks for your suggestions. We apologize for the incomplete literature review due to our incomplete survey, and we have actively reviewed the relevant papers and added the missing literature to the introduction. In addition, the contributions of this paper are highlighted in a targeted manner in the introduction. The main contributions of the paper are: we divide the ARM algorithms according to different pattern types and analyze the respective evaluation metrics. In addition, we compared each of them from three perspectives: biologically inspired, physically inspired, and other algorithms, and provided a paragraph describing the algorithms and a summary of their features in each group. We expect that this will provide some reference for research scholars in this field.

Point 3: Please read the paper carefully for English language style and accuracy, and make appropriate corrections and changes. With assist of professional experts rectify the grammatical and vocabulary error in the paper wherever necessary

Response 3: Thanks for your suggestions. We regret any grammatical or expressive errors in the English language. The grammar and English expressions in the text have recently been reviewed and rectified, and we will continue to improve our English study in our future studies and research to lessen the likelihood of such issues.

Point 4: The manuscript has sufficient results, but the final findings should necessarily be validated and compared with the results of standard diesel fuelled engine. Further, the study should clearly indicate the improvement or decline in results. At the end of each section in the results, authors should present concluding remarks with critical discussion in their own words.

Response 4: Thanks for your suggestions. All your comments are very important. I found the shortcomings in my work, and I will follow your suggestions to improve my research and achieve more in the future.

Point 5: The real applications and meaning of this study should be further elaborated in the section of Conclusion. It should only contain the significant outcomes and the take-home message and novel findings of this manuscript. Authors are advised to rewrite the conclusion section and give some insights about the variations in engine output characteristics numerically.

Response 5: Thanks for your suggestions. We have seriously considered your comments and completed the relevant changes . This paper primarily analyzes the research status of ARM from the standpoint of the PSO algorithm, and it summarizes and composes the ARM from three perspectives: mode types, evaluation metrics, and comparison with other algorithms. The analytical results of the PSO algorithm are shown in Figure 5, and the future outlook of PSO can be found in Section 6.

Author Response File: Author Response.pdf

Reviewer 3 Report

Overall, this manuscript seems to be a qualified review paper, and the literature and relevant techniques investigated are relatively detailed. However, I think the format of this manuscript should be improved, such as:

(i) The resolution of the figures needs to be improved, and some parts of Fig. 1 and Fig. 8 can not be seen clearly even if they are enlarged;

(ii) There are too many sub sections, and the font of different section titles and serial number patterns are too complex to confuse readers. It is suggested that the author optimize the manuscript structure;

(iii) The comma in the formula embedded in the text is obviously different from the comma in the text, which is not appropriate. Besides, even in the formula, there should be a space between the comma and the subsequent letters. So the authors are recommended to check and modify these problems.

(iv) I think the initial letter of where does not need to be capitalized when explaining the formula. It is suggested that the author check and modify it.

Author Response

Response to Reviewer 3 Comments:

Point 1: The resolution of the figures needs to be improved, and some parts of Fig. 1 and Fig. 8 can not be seen clearly even if they are enlarged.

Response 1: Thanks for your suggestions. I have inserted the images into the paper in ".vsdx" format to ensure that the images are still clear when enlarged.

Point 2: There are too many sub sections, and the font of different section titles and serial number patterns are too complex to confuse readers. It is suggested that the author optimize the manuscript structure.

Response 2: Thanks for your suggestions. To further optimize the manuscript structure, we merged some sections and reduced some of the sub-sections. At the same time, we have simplified the font of different section titles and serial number patterns to avoid confusing readers.

Point 3: The comma in the formula embedded in the text is obviously different from the comma in the text, which is not appropriate. Besides, even in the formula, there should be a space between the comma and the subsequent letters. So the authors are recommended to check and modify these problems.

Response 3: Thanks for your suggestions. In the article, we used MathType software to enter the formulae, in which we set the font to "Palatino Linotype,"  the same as the journaling font, to ensure that the formulae and the paper's commas are consistent. Besides, we have also added spaces between the commas and the subsequent letters in light of your suggestions.

Point 4: I think the initial letter of where does not need to be capitalized when explaining the formula. It is suggested that the author check and modify it.

Response 4: Thanks for your suggestions. We checked and modified the formula in the full text and changed the initial letter of where to lowercase.

Author Response File: Author Response.pdf

Reviewer 4 Report

1.  The Tabu search method in Figure 1 should be explained in detail.

2.  The  symbol apha from Equation (1) to Equation (3) should be explained in detail.

3.  More tested examples and explanations should be listed in section 4.

4.  The more introductions in Association Rule Mining (ARM) should be listed in References.

5,  The search method using convolution neural network(CNN) should be added into introduction, 

Author Response

Response to Reviewer 4 Comments:

Point 1: The Tabu search method in Figure 1 should be explained in detail.

Response 1: Thanks for your suggestions. We have added a detailed description of Figure 1 in the introduction of the article, which can be found in lines 70–79 of the article. The contents of this section are as follows:

In order to analyze these papers more systematically and comprehensively, we divide the ARM algorithms according to different pattern types. They are fuzzy association rule mining, rare association rule mining, numerical association rule mining, quantitative association rule mining, binary association rule mining, high-utility association rule mining, and classification association rule mining, and they are shown in blue in the inner circle of Figure 1. In addition, we compare them from three perspectives: biologically inspired, physically inspired, and other algorithms, and list the key information of each algorithm under the different perspectives, as shown in the outer circle of Figure 1. The specific description and summary of the characteristics of each algorithm can be found in Chapter 2 of the article.

Point 2: The symbol alpha from Equation (1) to Equation (3) should be explained in detail.

Response 2: Thanks for your suggestions. I have added the explanation of alpha from Equation(1) to Equation(3) in light of your suggestion. indicates the membership grade ofin the k^th records.

Point 3: More tested examples and explanations should be listed in section 4.

Response 3: Thanks for your suggestions. In Section 4, we add more detailed information about the PSO-based ARM algorithm, including fitness value calculation, population generation, and search for the best particle. Among them, in the process of finding the best particle, we also need to set the boundary conditions and termination conditions. The details can be found in Section 4.1.4.

Point 4: The more introductions in Association Rule Mining (ARM) should be listed in References.

Response 4: Thanks for your suggestions. We strongly agree with your comments. We have actively searched for papers related to association rule mining and added them to the introduction section of the article, which we hope will make the article more complete.

Point 5: The search method using convolution neural network(CNN) should be added into introduction.

Response 5: Thanks for your suggestions. In the introduction, we have added the following papers:

https://ieeexplore.ieee.org/abstract/document/9686436/

https://www.hindawi.com/journals/cin/2022/4851615/

https://link.springer.com/article/10.1007/s13369-022-06870-x

Thanks again for your advice, he is very important. Thanks to your advice, I have discovered my current shortcomings in my work. I will follow your advice in the future to improve my research and achieve more.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Reviewer 3 Report

Thanks for addressing my comments. I think this version could be publishable. 

Reviewer 4 Report

This paper has been revised according to the reviewer's comments. This paper should be accepted.