Cooperative Attention-Based Learning between Diverse Data Sources

Round 1

Reviewer 1 Report (Previous Reviewer 2)

Although authors have revised the paper most of the concerns and shortcomings are still present.

The description of the model is still unclear. Equations are poorly described, no examples are provided. Some pictures are provided with no caption and are not cited trough the text. Some images have a very poor resolution and are not easy to follow (figs. 3 and 4). The paper has two figures 8. The objective of the paper is not properly defined. The experimental analysis looks not sound. Which are the metrics you use? How you measure the quality of your results?

Author Response

Second Review

Journal: Algorithms Manuscript ID: algorithms

Paper Title: Cooperative Attention Based-Learning Between Diverse Data Sources
Authors: Harshit Srivastava and Ravi Sankar

Dear Reviewers,

We greatly appreciate the constructive feedback provided by the reviewers, which has guided us in refining and enhancing our manuscript. In response to their comments, we have carried out a thorough update of the paper, which includes the clarity in objectives, as well as the reorganization of its structure to ensure a more coherent presentation of our findings. To further improve clarity, we have judiciously added and removed images and results, considering the reviewers' suggestions.

Additionally, we have streamlined the text by removing extraneous content and incorporating relevant performance metrics to facilitate a more comprehensive understanding of our work. In order to simplify Figure 1, we have reduced the number of connections, thus making it more accessible to our readers. Lastly, to maintain the flow of the paper, we have shifted certain content to the appendix. We believe that these revisions have significantly strengthened our paper and addressed the concerns raised by the reviewers. Below, you will find a detailed, comment-by-comment response to each of the reviewers' remarks, further elaborating on the changes we have made.

Reviewer #2 Comments and Authors Reponses:

• The description of the model is still unclear.

[Response]: We appreciate the reviewer's feedback on the clarity of the model description. In response to this comment, we have revised the model description and added more detail to ensure that the model is well-defined and easy to understand. Specifically, we have provided a comprehensive description of the cooperative learning-based multi-agent framework for disease spread analysis, including the observation model, the spread variable determination, and the control algorithm.

In addition to these revisions, we have made structure changes and written the methodology section in more detail, offering a better explanation of the processes involved. To further enhance comprehension, we have included a redrawn figure 3 diagram to visually represent the various steps in the model, as well as a Figure 7 and Table 2 summarizing the key metrics used to measure the quality of the results.

We believe that these revisions will help clarify any confusion and provide a better understanding of the model for our readers.

• Equations are poorly described, no examples are provided.

[Response]: Thank you for the feedback. We apologize for any confusion caused by the lack of examples and unclear description of the equations in our paper. We understand the importance of providing a clear and comprehensive explanation of our model and its mathematical components.

To address this issue, we have updated our paper to include more detailed and explicit explanations of the equations, along with few relevant examples to better illustrate their applications. We hope that this will provide a clearer understanding of our model and its mathematical foundations.

Additionally, we are happy to provide further clarification or examples upon request, and we welcome any specific feedback or questions regarding our mathematical methods. Thank you for bringing this to our attention, and we appreciate the opportunity to improve the clarity and comprehensiveness of our paper.

• Some pictures are provided with no caption and are not cited trough the text.

[Response]: Thank you for your valuable feedback. We apologize for the oversight in not providing captions for the figures in the paper. We made sure to update the paper with proper captions and references throughout the text. We believe that this will significantly improve the clarity of the paper and make it easier for readers to follow along with the presented material. Additionally, we will make sure to include proper citations for all figures in the text. Thank you again for your feedback and for helping us improve the quality of our paper.

• Some images have a very poor resolution and are not easy to follow (figs. 3 and 4).

[Response]: We appreciate the reviewer's feedback and agree that the quality of the images in Figs. 3 and 4 could be improved. In response to this feedback, we have re-generated these figures with higher resolution and improved clarity to better facilitate understanding. We have also included appropriate captions and cited these figures throughout the text to ensure that they are more effectively integrated into the manuscript. Thank you for bringing this issue to our attention and helping us to improve the quality of our work.

• The paper has two figures 8.

[Response]: Thank you for pointing out this discrepancy in our manuscript. In response to Comment 5, we have resolved the issue by carefully examining the figures and correcting the numbering. We apologize for any confusion this may have caused and appreciate your diligence in identifying the error.

• The objective of the paper is not properly defined.

[Response]: Thank you for your comment regarding the objective of the paper. We appreciate your feedback and will address this issue accordingly.

We understand that the objective of the paper was not explicitly defined and have taken steps to clarify it in our revised manuscript. We have rewritten the objective in detail and created a dedicated section for it before the methodology.

We hope that our revised manuscript provides a clearer understanding of our objective and the approach we have taken to achieve it. We appreciate your feedback and look forward to hearing from you regarding the revised version of our manuscript.

• The experimental analysis looks not sound. Which are the metrics you use? How you measure the quality of your results?

[Response]: Thank you for your comment regarding the experimental analysis in our paper. We used several metrics to evaluate the performance of our proposed model, including accuracy, precision, recall, F1 score, and confusion matrix. We measure the quality of our results by comparing them with the ground truth data and evaluating the model's performance in predicting the spread of the disease accurately.

We conducted analysis on a dataset of COVID-19 cases in the United States to evaluate the effectiveness of our proposed model. The results showed that our model can accurately predict the spread of the disease with an overall accuracy of 67%. We also conducted sensitivity analysis to evaluate the impact of different parameters on the model's performance.

We apologize for any confusion regarding the experimental analysis in our paper and will make sure to provide more detailed explanations and results in future revisions.

Author Response File: Author Response.pdf

Reviewer 2 Report (Previous Reviewer 1)

I thank the authors for their work; basically, as I mentioned earlier, the main issue of this work is its presentation way, which is very poor. Indeed the presentation has been improved, but I also noticed authors had added more text. Part of our responsibilities as researchers is to make to work readable even to the community.

Anyway, I provided my comments which suggest another major revision for this paper, specifically in figures and figure legends, and actually carried steps. I saw the published article in 2021, and it was much more organized than this one.

Comments for author File: Comments.pdf

Author Response

Second Review

Journal: Algorithms Manuscript ID: algorithms 2252205

Paper Title: Cooperative Attention Based-Learning Between Diverse Data Sources
Authors: Harshit Srivastava and Ravi Sankar

Dear Reviewers,

We greatly appreciate the constructive feedback provided by the reviewers, which has guided us in refining and enhancing our manuscript. In response to their comments, we have carried out a thorough update of the paper, which includes the clarity in objectives, as well as the reorganization of its structure to ensure a more coherent presentation of our findings. To further improve clarity, we have judiciously added and removed images and results, considering the reviewers' suggestions.

Additionally, we have streamlined the text by removing extraneous content and incorporating relevant performance metrics to facilitate a more comprehensive understanding of our work. In order to simplify Figure 1, we have reduced the number of connections, thus making it more accessible to our readers. Lastly, to maintain the flow of the paper, we have shifted certain content to the appendix. We believe that these revisions have significantly strengthened our paper and addressed the concerns raised by the reviewers. Below, you will find a detailed, comment-by-comment response to each of the reviewers' remarks, further elaborating on the changes we have made.

Reviewer #2 Comments and Authors Reponses:

• Figure 1. still not fixed, there are white arrows covering some texts such as Strategy Creation. The quality of the figure must be improved.

[Response]: Thank you for your valuable feedback. In response to the feedback on Figure 1, we have carefully reviewed and updated the illustration. Our revisions involved redrawing the figure to reduce complexity and ensure that no essential information is omitted. We believe that these modifications have made the figure more accessible and comprehensible for our readers, effectively addressing the concerns raised. In addition, we have added a caption to the figure, explaining the different components and their relationships within the model. We hope that these changes will improve the clarity of the figure and enhance the overall readability of the paper.

• Figure 3. needs to be sharper; many texts are not readable:

[Response]: Thank you for your valuable feedback on our paper. We appreciate your comments and would like to address the issue you raised regarding Figure 3.

We apologize for the unclear and unreadable texts in the figure. We understand the importance of visual aids in scientific papers and will take the necessary steps to improve the clarity of our figures for better readability. Specifically, we have enhanced the resolution and clarity of the image so that all texts and labels are legible. Additionally, we will ensure that all figures are properly captioned and cited within the text.

Thank you again for your feedback, and we hope that these improvements will enhance the quality of our paper.

• Figure 8 legend is wrong. Figure 8a was mentioned twice. Please redo.

[Response]: We appreciate your attention to detail in identifying the issue with Figure 8's legend and the repeated mention of Figure 8a. We understand the problem and are grateful for your guidance. In response, we have re-plotted the results, taking care to correct the legend and ensure that each figure part is accurately referenced. Furthermore, we have enhanced the revised figure by including additional information and key metrics to provide a more comprehensive representation of our results. We believe that these improvements will address the concerns raised and contribute to a clearer understanding of our findings.

• I saw the authors mention the figure legends in the text. In my previous comment, I mentioned making the figure legends below each figure, so I reduced the text size and made the figures more informative. For example, “Figure 2 is a tree-like diagram that illustrates how the agents can make decisions and accumulate rewards based on their actions and the state of the system… /// this can also be as figure legend, not just one sentence. The authors, in their replies, mentioned, “We have updated the figures and legends with descriptions in the figures”, this is not true, I can only see they changed the color of the text…

[Response]: Thank you for your feedback regarding the figure legends. We apologize for any confusion or misunderstanding. Upon further review, we agree that the legends could have been more informative and clearer. We have updated the figure legends as per your suggestion and added more detailed descriptions to each figure to make it easier for readers to understand the content of the figures. We also ensured that all figures are properly cited in the text to provide a clear connection between the text and the figures. We appreciate your valuable feedback and hope that these changes improve the readability and clarity of our paper.

• Please combine the subtitles related to the methodology under the method section

[Response]: Thank you for your feedback. We have revised the manuscript by combining the subtitles related to the methodology under the method section to provide a more organized and coherent structure. The revised manuscript now presents the methodology in a step-by-step manner, starting with the observation model, followed by the cooperative learning and strategy creation, spread-based analysis for cooperative learning, and finally the determination of the spread dynamic variable. We believe that this reorganization of the manuscript will enhance the clarity and readability of the methodology section. Once again, thank you for your valuable feedback.

• Lines 605 to 609 and 712 and 716 contain multiple repetitive data.

[Response]: We appreciate the reviewer's comment on the repetitive data in our manuscript. Upon reviewing the mentioned lines, we agree that there is some repetitive information. We apologize for any confusion that may have arisen due to this error. In order to address this issue, we will revise those lines and streamline the information presented in those sections, ensuring that the information is clear and concise. We thank the reviewer for bringing this to our attention and we will make the necessary changes to improve the quality of our manuscript.

• Lines 586: This can be seen in the first column of the table, … which table ? of course, you mean table 2, but you forgot to add it in the new submission.

[Response]: We apologize for the oversight in not including Table 2 in the revised submission. We have made the necessary adjustments and added Table 2 to the manuscript. Thank you for bringing this to our attention and allowing us the opportunity to correct our mistake.

• I understand that you mentioned there is no specific methodology. Would it be good to make another figure showing the steps you performed instead of bringing a theoretical model? For example, in figure 1 (which is good but needs improvement), you mentioned only theories, for example, you said “Social Data”, would it be fine if you create a figure showing the actual work that was done, such as “Twitter tweet, N= 185,755”, etc.

[Response]: Thank you for your valuable feedback. We agree that a more detailed description of the methodology and steps involved in the study would be helpful for readers. To address this concern, we have updated our paper with that shows the steps we took in our study, including the data collection process and the analysis steps and how we verified it the results.

We provide few details of the data collection process and referenced it in the manuscript, which includes the collection of tweets from the Twitter API. We have also included the number of tweets we collected (N=185,755) and the time period over which we collected the data. Additionally, we have included the steps we took to preprocess the data and filter out irrelevant tweets, including retweets and non-English tweets.

Furthermore, we have updated the caption for Figure 1 to provide more context and to cite it appropriately throughout the text. We appreciate your feedback and hope that these updates will provide readers with a clearer understanding of the methodology used in our study.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report (Previous Reviewer 1)

I thank the authors for the complete revamp of the paper, for the explanations, and for adding new figures. I can also see that the introduction and literature review became even longer.. but the authors required this to explain the importance of their model and overall theory. I also think that the literature review part can increase the impact of this paper to be a reference for upcoming, similar articles. 

This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.

Round 1

Reviewer 1 Report

The paper entitled “Cooperative Attention Based-learning Between Diverse Data Sources”. The authors proposed a new approach to aid in determining the spear of COVID 19 pandemic across US.

Major concerns.

1.       The paper was plagiarized with 83% and uploaded to the University of South Florida. The authors must contact the editor and clarify this issue.

2.       As the paper was previously submitted as a student paper, I can see that the authors neither followed the criteria of the journal nor used the journal template.. also, the reference style is not correct. Please refer to the journal reference style.

3.       Also I am not sure what is the clinical usage of this methodology at the moment, especially you said you had to receive data from the hospital to apply this algorithm to it. So, in the real world,  If we provide you with hospital data, you should (according to your paper) go back to the time and check the social media to align the information with it and build us a module that predicts the spread rate of the virus with 67% accuracy. Please highlight how many hours this work will take and if it is practically possible. This is missing in your entire manuscript. What are the applications, the actual ones... Therefore, you can add a new figure showing the steps of an author/reader how to implement your algorithm. 

4.       Figure 1. On which basis you determined the four stages and the figure need to be polished? Some arrows are not aligned correctly, such as those in environment determination. Also, the legend of the figure needs to be the lines 90-96, not only Decision Spread Model Block Diagram. Also, do you assume its location is proper? Was it all original?

5.       Line 86: This all sector for an objective is redundant. Move it without a header to the introduction.

6.       Line 159, reference 41, needs to be included with 34-37.

7.       Line 178-179, here you again started to write about the objectives…

8.       Line 97: you did a great job in the literature review, and I found some repetitive data in the introduction. Please revamp these two sections and ensure a logical flow of the data.

9.       Line 180: this section is considered the core of your paper, and it is unclear if you created these equations 1-7 or adapted it from other literature.

10.   The methodology section is vague and not clear. Make sure to have a heading calling methodology and highlight subheadings for each variable. Also, it is not visible how you reached the results; for example, using special software, language packages are missing. You can check other manuscripts in the journal and check how they are writing their manuscript.

11.   Please add a section for the limitation and opportunities for research

12. Photos need to be sharper

13. important equations can stay in the manuscript; others can be in the supplementary

14. Do you think 67% accuracy is good for such a model, especially if we are talking about spreading (360 degrees)? 

15. for all figures, make sure to make the figure legend descriptive as much as you can and delete the repetitive information from the manuscript body

Reviewer 2 Report

In this paper authors propose e model based on a cooperative strategy which join different data sources of social network and physical network to understand disease spread. The problem is certainly interesting, however the paper presents several major issues:

1.The theoretical description of the model is poorly presented and it has to be completely restructured. Most of the equations use a notation hard to follow and these sometimes these result unsound. Authors first have to define, through a table all the notation used through the paper (at line 369 is present a table which is not referenced through the paper, such a table could be a starting point but has to be moved at the beginning and extended with all the symbols used). Each equation has to be  carefully revised and clarified also with examples. Sometimes the notation looks not clear (i.e. equations with summation symbol with different index t,n what does it mean? Or line 311 what is lnx_n? a parenthesis is missing).

2. Most of the pictures used through the paper are not properly described and result unclear. Also these have a resolution very low.

3. The experimental analysis is not clear at all. First the goal of the analysis has to be clearly stated: the prediction of the relation between spreading of information on a topic (covid19) and the spreading of disease. Second a clear description of the data used has to be provided. Also a repository with data should be made available. How the grouping of tweets has been done? Action influence, spread influence and location influence are not clear. Figure 7, 9 and 10 are mostly unreadable. Which are the x axes of figures 7 and 9 ? Is the code used to develop the model available?   The results of table 2 have to bee properly described.