Predictive Analytics for Sustainable E-Learning: Tracking Student Behaviors

Round 1

Reviewer 1 Report

This paper proposed a pipeline for recognizing students' unauthorized behaviors during online learning using recording videos. I have several questions concerning this research:

1. In terms of the technical part, the advantages of the proposed pipeline are not clearly stated. For example, the SOTA algorithms like YOLO/SSD can perfectly handle object detection task, and other out-of-the-box methods for instance segmentation or skeleton extraction. Are the videos processed frame by frame or using some inter-frame relations?

2. For results part, actions like Sad and Angry in Figure 11 are not exist in Table 1, why? And the bar in Figure 11 and 12 are not green bar (mentioned in Ln 300).

3. The authors only demonstrate the results of HAR (Human Action Recognition), and the discussions on the results/relations to education (improvements on effects of online learning) and sustainability are not discussed.

4. For the reference section, 42 out of 46 references are the publications of the last author of the manuscript.

5. Missing Section 4.1.

6. The relationship between outdoor cases (Figure 9) and online learning? The first two examples in Figure 10 are the same type.

7. The table on Page 4 is not numbered. And why references 12-20 are described using text while references 21-25 using table?

8. Equation (4) in Ln 291 should be Equation (8).

Extensive editing of English language are required. Also, inconsistent expression exits. For example, text in Ln 91-95 is not the fact the paper organized.

Author Response

Thanks for your valuable comments. 

Answers are provided in attached file. 

Please consider.

Author Response File: Author Response.pdf

Reviewer 2 Report

Please describe the population, sample and data time period. Please describe in more detail figure.8 how to solve the problem and describe more detailed result data interpretation.

Author Response

Thanks for your valuable comments. 

Answers are provided in attached file. 

Please consider.

Author Response File: Author Response.pdf

Reviewer 3 Report

The topic of this study titled "Predictive Analytics for Sustainable E-Learning: Tracking Student Behaviors" is interesting and attractive. However, this work still has some problems that should be improved, as follows.

1) The abstract should be rewritten to highlight the main purpose of this work.

2) Section 1 should be rewritten relying on some aspects of:

- Why is e-learning timely? you need to present some COVID-19 figures to clear the topical.

- E-learning/smart learning, etc., are deployed on IoT and 5G/6G infrastructure, hence, you should indicate the relationship of e-learning, 5G/6G in the IoT era. Please refer to: 

a) Innovative Trends in the 6G Era: A Comprehensive Survey of Architecture, Applications, Technologies, and Challenges, IEEE Access.

3) All Equations should be reformatted

4) All Figures should be enhanced with a minimal resolution of 300 dpi.

5) In Section 5, Discussion, the authors should mention emerging technologies such as AI, Digital Transformation, IoT, and Edge Computing to drive e-learning. Please refer to:

b) AI and Digital Transformation in Higher Education: Vision and Approach of a Specific University in Vietnam, Sustainability

6) References [1-9] in the period [2012-2013] are old too. They should be replaced by the newest references.

Moderate editing of English language required.

Author Response

Thanks for your valuable comments. 

Answers are provided in attached file. 

Please consider.

Author Response File: Author Response.pdf

Reviewer 4 Report

Title: Predictive Analytics for Sustainable E-Learning: Tracking Stu- 2 dent Behavior.

Here are my comments:

 

1.    Abstract needs improvement by show the methodology more clearly.

2.    The authors should clearly state what are the problems, objectives, contributions and the novel aspects of the paper in more details. 

3.    In section 2 related work, the authors show summary of some studies, the table should have a title and number.

4.    In addition, the same table needs to present more studies showing the year of publication, advantages, and disadvantages of these studies.

5.    The related works should provide and discuss in detail the differences between the proposed method and other methods in the literature, to show the gap of the proposed solution with existing works.

6.    The proposed method is the core of the paper. I expected to see the details of the proposed method as it is the most important.  I recommend authors to improve this section.

7.    It will be great to have some scheme, Pseudocode, algorithm steps …etc of the proposed method to visualize the steps, relationship of each step and the role of each step clearly.

8.    Authors should show the advantages and disadvantages of the proposed solution. In addition, please explain clearly what the exact significance of the proposed method to the field of study.

 

 

 

Thank you for the work.

 

Minor editing of English language required, some grammatical errors should be fixed. 

Author Response

Thanks for your valuable comments. 

Answers are provided in attached file. 

Please consider.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Thank you for the efforts on revising this paper. Please add more explanation on the reason choosing Viola-Jones algorithm rather than DL based methods in the paper.

The overall writing should be polished before publication.

Author Response

Thanks for your valuable comments. 

Solutions are attached. Please consider.

Author Response File: Author Response.docx

Reviewer 2 Report

-

Author Response

Thanks for your valuable comments. 

Solutions are attached. Please consider.

Author Response File: Author Response.docx

Reviewer 3 Report

Although the revised version is improved significantly compared to the original version. However, the presentation quality is very low. 

- The authors must be enhanced the minimal resolution of the 300 dpi.

- The authors should re-check all comments from reviewers double-check.

Moderate editing of English language required.

Author Response

Thanks for your valuable comments. 

Solutions are attached. Please consider.

Author Response File: Author Response.docx

Reviewer 4 Report

The authors addressed some of my comments. The paper now looks better and more professional, I feel some improvement has been made. However, referring to my previous comments, there are some important points not revised clearly, correctly and some points not answered properly. Therefore, here I have some comments from my previous comments that should be addressed more clearly by the authors: 

Comments 3 and 4, authors ignored these comments as removed the table. The benefit of using table, the table can summarize many studies showing the year of publication, advantages, and disadvantages of these studies. The readers can easily detect the related studies with all details they needed.  

 

Comment number 6, authors answered as “they have been revised it”.it is still not improved to be clear for the reader.

Comment number 7, Algorithm 1, not clear as a text, it needs to improve the clearance of the text instead of show it as image.

Comment 8, the authors try to improve it and they schussed so far, but this point should be discuss in more detail to show the significance of the proposed solution and the role the predictive analytics for sustainable e-Learning.

 

 

Thank you for the work.

Minor editing of English language required, some grammatical errors should be fixed.

Author Response

Thanks for your valuable comments. 

Solutions are attached. Please consider.

Author Response File: Author Response.docx

Round 3

Reviewer 3 Report

Although, this version has been revised, however, there are many errors, for example:

1) The title of Algorithm 1 in line 142 is incorrect, "Algo: Multi-stage processing and Classification pipeline for Behavior Analysis."

2) The presentation Algorithm 1 must be revised.

Minor editing of English language required.

Author Response

Thanks for your valuable comments.

Solutions are attached in the attached file. 

Please consider.

Author Response File: Author Response.docx