Course Prophet: A System for Predicting Course Failures with Machine Learning: A Numerical Methods Case Study
Round 1
Reviewer 1 Report
The paper titled " Course Prophet: A Machine Learning-Based System for Early Prediction of Student Failure in Numerical Methods Course in the Bachelor’s degree in Engineering at the University of Córdoba, Colombia" is reviewed. A number of related recommendations are presented as follows:
· “In another two studies, the prediction is performed before the course begins, using the grades achieved by the student in prerequisite courses” in this sentences what The author/(s) mean? Which two studies? Can the studies be the authors of the cited studies?
· In the introduction part, it may be better to make general definitions instead of the study purpose.
· The Discussion, which are stated in the form of items, are mentioned as the findings. Interpretatively, what the results revealed could be conveyed.
Author Response
Dear referee, first of all, thank you for your comments, these were insightful in improving the manuscript.
Comment 1:
In another two studies, the prediction is performed before the course begins, using the grades achieved by the student in prerequisite courses” in this sentences what The author/(s) mean? Which two studies? Can the studies be the authors of the cited studies?
Correction:
In contrast to the studies mentioned earlier, in [3,6] the goal is to forecast students’ risk of failure prior to the course beginning, using their grades in prerequisite courses. In [3], grades are represented as ordinal data rather than quantitative data, and students’ ages are also used for prediction in addition to grades. In [6], the predictive input variables encompass scores in the admission test and grades in prerequisite courses. Notably, this latter work is dedicated to the Numerical Methods course within the Bachelor’s degree program of Systems Engineering at the University of Córdoba in Colombia. In the present paper, we have achieved enhanced results in comparison to the findings published in [6].
Comment 2
In the introduction part, it may be better to make general definitions instead of the study purpose.
correction
I included the definitions required to understand the background in the introduction.
Comment 3
The Discussion, which are stated in the form of items, are mentioned as the findings. Interpretatively, what the results revealed could be conveyed.
Correction
I removed the items, and outlined the discussion in a different way, interpreting the results of the evaluation.
Reviewer 2 Report
Titles can be made shorter and more interesting
As the research question is more useful in the introduction, the abstract should only state the purpose of the research, not the research question.
In the preface, the aim of the study was "designing an intelligence system that predicts if a given student is at-risk of failing the Numerical Methods course before the lectures start, based on the student’s performance in prerequisite courses" and in the methods, the aims were "to predict the risk of course failure for students, and to achieve this, we adopt a quantitative approach using machine learning methods in lieu of solely relying on statistical approach"
What exactly will be the subject matter of this article?
Tables and graphs should not just be presented without any explanation of how to read them but should be accompanied by an explanation of their meaning.
The conclusion should directly present the state of the research, the novelty of this research and what is the final conclusion of the discussion that has been carried out. There is no need to repeat the aims, methods and research process.
Sentences or explanations are often repeated. The article should be brief, concise and clear. However, it should also be detailed in its discussion.
Except for concepts that are constant and stable, it is worth reconsidering the use of outdated references, especially those related to the learning process.
Comments for author File: Comments.pdf
Author Response
Dear referee, I deeply appreciate your comments, these were insightful and helped me to improve the quality of the manuscript
Comment 1
Titles can be made shorter and more interesting
Correction:
I changed the title, now it is: Course Prophet: A System for Predicting Course Failures with Machine Learning: A Numerical Methods Case Study
Comment 2:
As the research question is more useful in the introduction, the abstract should only state the purpose of the research, not the research question
Correction
I fixed this, by changing the abstract and introduction according to your instructions.
Comment 3:
In the preface, the aim of the study was "designing an intelligence system that predicts if a given student is at-risk of failing the Numerical Methods course before the lectures start, based on the student’s performance in prerequisite courses" and in the methods, the aims were "to predict the risk of course failure for students, and to achieve this, we adopt a quantitative approach using machine learning methods in lieu of solely relying on statistical approach"
What exactly will be the subject matter of this article?
Correction:
I modified the introduction in order to keep the coherence of the manuscript
Comment 4
Tables and graphs should not just be presented without any explanation of how to read them but should be accompanied by an explanation of their meaning
Correction:
I've expanded the explanation in the captions of the figures and table legends to lead the reader to understand the presented information.
Comment 5
The conclusion should directly present the state of the research, the novelty of this research and what is the final conclusion of the discussion that has been carried out. There is no need to repeat the aims, methods and research process.
correction
I wrote the conclusion again, presenting the state of the research, and explaining the novelty of the study and the conclusion of the discussion. I removed the summary of the paper, including goals, methods, and research process.
Comment 6
Sentences or explanations are often repeated. The article should be brief, concise and clear. However, it should also be detailed in its discussion.
Correction:
I systematically removed every repetition I found. You're utterly right, the manuscript was full of redundant explanations and arguments making the reading daunting, exhausting, and boring. Thanks for this comment.
Comment 7
Except for concepts that are constant and stable, it is worth reconsidering the use of outdated references, especially those related to the learning process
Correction
While I keep the references to the classical machine learning methods adopted in the study, the references to related work are updated as possible.
Reviewer 3 Report
The paper entitled "Course Prophet: A Machine Learning-Based System for Early Prediction of Student Failure in Numerical Methods Course in the Bachelor’s degree in Engineering at the University of Córdoba, Colombia" investigated several statistical and machine learning methods to provide a prediction for students failure in a specific university course based solely on their performance in the prerequisite courses. The methods are explained concisely, and the results and discussion provided necessary statistical evaluation to find the best machine learning method to address this problem. However, the development of Course Prophet has not been elaborated in the paper. It is suggested that the author(s) explain about the development of the Course Prophet system to complete the results and discussion section. This paper is written in a proper academic English, however minor revisions are needed to enhance the quality overall.
Comments for author File: Comments.pdf
Please revise the English according to the notes/comments from reviewer in the PDF file.
Author Response
Dear referee, thanks for your comments, with your help I've improved the manuscript.
comment:
the development of Course Prophet has not been elaborated in the paper. It is suggested that the author(s) explain about the development of the Course Prophet system to complete the results and discussion section. This paper is written in a proper academic English, however minor revisions are needed to enhance the quality overall.
correction:
I included a section where I delve into the details of the Course Prophet components, and I've centered the discussion on this subject. Finally, I've reviewed and rewritten almost every part of the article to remove typos and verbose descriptions. Thanks for letting me notice this problem.
Reviewer 4 Report
Authors developed a new prototype to predict the risk of course failure based on student’s achievements.
I didn’t understand the second contribution? An intelligence system prototype was developed or utilized? It was not clearly stated! Did you develop a novel classification model by modifying Gaussian processes or did you just utilize it?
This paper looks like more review paper than research one, because a novel dataset was created which can be applied to state-of-art classification models, but I didn’t see any novel model proposed by authors.
In experimental evaluation part, you can also include rule-based algorithms such as PART, RIPPER into your comparison, because those algorithms perform well on classification datasets. Did you check the Naïve Bayes model?
This paper should be proofread by English native speaker!
Here are some minor revisions:
1. In line 85: “age”?.
2. In lines 110-112 and 116-118: There is a repetition: “Can an artificial intelligence system effectively learn patterns in students’ academic history to predict whether a given student is at risk of failing a course, based only on their performance in prerequisite courses? “and “our goal in this study was to investigate whether an artificial intelligence system can effectively predict course failure risk solely based on students’ performance in prerequisite courses”. Merge these two sentences!
3. In line 146: Why ? Shouldn’t be
?
4. In line 150: “is the worst that a given student achieved”, word “grade” is missing? Correct it in next lines as well.
5. In line 218: “understands” should be corrected as “understand”.
6. In line 300: Equation numbers are missing.
Comments for author File: Comments.pdf
This paper should be proofread by English native speaker!
Author Response
Dear referee, I thank you a lot for your valuable comments. These were insightful to improve the quality of the article.
Comment 1
I didn’t understand the second contribution? An intelligence system prototype was developed or utilized? It was not clearly stated! Did you develop a novel classification model by modifying Gaussian processes or did you just utilize it?
correction
I didn't develop a novel classification model, to clarify this point I've rewritten some parts of the manuscript and written a new section in it.
Comment 2
This paper looks like more review paper than research one, because a novel dataset was created which can be applied to state-of-art classification models, but I didn’t see any novel model proposed by authors.
Correction:
You made me realize I didn't explain the novelty of this study, thereby, I've re-written several parts of the paper to communicate it to the reader.
Comment 3
In experimental evaluation part, you can also include rule-based algorithms such as PART, RIPPER into your comparison, because those algorithms perform well on classification datasets. Did you check the Naïve Bayes model?
Correction
I wrote this as a direction for further work:
Furthermore, delving into the possibility of extending Course Prophet into a recommender system for course selection, tailored to each student's performance, holds significant promise. By offering personalized course recommendations that align with their academic strengths and requirements, we can substantially enrich the educational journey of students. In pursuit of this goal, we plan to investigate unsupervised learning techniques, including but not limited to association rules, dimensionality reduction, clustering, and so forth. These explorations aim to further refine Course Prophet's capabilities and enhance its utility in guiding students toward successful academic paths.
Comment 4:
This paper should be proofread by English native speaker!
Correction:
I've verified and corrected every typo. Moreover, someone else with good proficiency in the English tongue read the manuscript to find other typos.
Comment 5:
In line 146: Why? Shouldn’t be?
Correction:
I clarified this point in the introduction and methodology sections
Comment 6
In line 150: “is the worst that a given student achieved”, word “grade” is missing? Correct it in next lines as well.
Correction:
fixed
Comment 7
In line 218: “understands” should be corrected as “understand”.
correction:
fixed
Comment 8
In line 300: Equation numbers are missing.
Correction
Fixed
Thanks, I have no idea why I left those equations references empty.
Round 2
Reviewer 2 Report
The authors have improved on the previous feedback to make this article suitable for publication.
In this section (Funding), it is necessary to add an explanation of the components in the table.
Many references are still old, it would be better to use the latest sources.
Comments for author File: Comments.pdf
Reviewer 4 Report
Revision is ok. Anyway please clearly state the contribution and novelty of the paper in abstract, introduction and conclusion