Using Artificial Intelligence to Predict Class Loyalty and Plagiarism in Students in an Online Blended Programming Course during the COVID-19 Pandemic
Round 1
Reviewer 1 Report
This article discusses an important, valuable and recent topic. The efforts of the authors are evident in completing this article. The authors have to carefully address the below concerns.
- Introduction Section: It is well written. However, we suggest dividing the contents of the introduction into subheadings to make this introduction simple and smooth for the reader.
- If the teacher prepared one of the questions from the lectures that all the students have, say 90% of the students answered that question with the same scores, duration time, and action logs. Are the parameters (scores, duration time, action logs ...) sufficient to check plagiarism without unfairness to students in terms of grades?
- English writing: This article requires extensive proofreading, there are a lot of typos and grammatical mistakes. The authors must carefully scrutinize the entirety of the article.
- List of references: It is recent, arranged in-text and sufficient. However, it requires minor improvement. Some search names in the reference list begin an uppercase letter for each word (such as [7], [8] ... etc.) and others use only an uppercase letter in the first word (such as [1], [2] … etc.), authors should standardize style. Journal names should be italic such as [11], [12] … etc. The authors should accurately scrutinize the list of references to eliminate all problems.
- Figures: Some Figures require redrawing to be clear in high resolution such as Figure 1, 7, 8, 9, 10, 11, 12, 13, 15 and 16. Figure 4 is not used in the text.
Author Response
Thanks for the comment. There are some answers to the questions.
- Thanks for the suggestion. We have modified add some sub-section titles in Introduction section. (Please see the content marked in red in Introduction Section)
- Yes. Because “Score” is one check item on the list, it only useful when the student copies the whole code as the same as the answer. Those students, who have plagiarism, just want to obtain high scores and pay lowest effort. From the observation, most of the students, who have plagiarism, only copy the code, paste the code to the system, and perform the judgement to obtain the score. It only generates 3 actions (copy, paste, judge) and the duration time is also short. That’s why we use the equation (4) to check. The ue is the time which we set to the 3 second in real case. In other words, no matter how fast the student types, it can’t be achieved if the students are normally programming. The same as equation (5), the ae is set to 3 (3 actions) in real case. Hence, the diligent students are not miss-classified to the students who have plagiarism if they are normally programming.
- The English of this paper is polished by native speaker.
- The references all been checked again. Thanks for the suggestions.
- The figures are redrawn. However, the figures such as the Figure 9 and Figure 11, are generated by using the program in 600 dpi already. I will try my best to redraw them again.
Author Response File: Author Response.pdf
Reviewer 2 Report
Dear authors,
I have identified the following aspects that from my point of view require your attention:
1) In the description of the identification data of the authors we have emails that are not institutional.
Please enter institutional email addresses if you do not work for Google LLC.
2) In line 20 we have the following sentence
"206 students are involved in the experiments"
A sentence does not start with the ordinal number written in numbers in English.
Please identify / document / highlight if the number of 206 subjects is significant in your research.
3) In line 27 to keywords I think it would be useful to have and
Convolution Neural Network, COVID ‐ 19 Epidemic
It is important to have representative keywords because when looking for potential readers to elaborate state of the art if we want to receive quotes we look for keywords and read the abstract first, followed by conclusions then read the article.
Here attention SARS-CoV-2 = COVID ‐ 19 is pandemic disease not epidemic disease!
4) From line 85 to line 87 please put in quotation marks because that phrase does not belong to you.
I have attached the similarity report.
5) The figures are generally not legible, please improve this aspect, possibly work in LaTeX.
6) In lines 281-283 an Object Tensor OT concept was defined. It is not clear to me the definition of this concept. Tensor calculus is not in the field of physics, because it applies there ... it's completely different. Tensor algebra and tensor calculus are a subdiscipline of mathematics.
Please make the correct connection to the concept defined by your Object tensor and TensorFlow in Python TM, if it exists.7) For figures 8 and 10 because I think you worked in Micro $ oft Word there is an Easy Code Formatter plugin that is free to highlight the code in C, which seems to be written in ANSI C.
Easy Code Formatter is free here:
https://appsource.microsoft.com/en-us/product/office/wa104382008?tab=overview
In the code in figure 8 and figure 10, however, something is missing after
# includes if the file extension is * .cpp in Windows ... variance or maybe
# includes "stdio.h" if the source extension is * .c on UNIX, Linux ...
Possibly for the second
# it includes
Preprocess directives are needed for inputs / outputs
such as conio.h, because the environment in which the source code was written, respectively the operating system, was not indicated.
8) For storing data and source code it is useful to use a repository that generates DOI code
A data source was identified in line 614
According to
https://www.mdpi.com/journal/data/instructions
The data source should contain a DOI number, as they can be subsequently cited.
There are countless free data warehouses, even at MDPI.
Comments for author File: Comments.pdf
Author Response
Thanks for the comment. There are some answers to the questions.
- Thanks for the suggestion. We have modified the email addresses.
- Thanks for the suggestion. We have modified the description.
- Thanks for the suggestion. We have modified the keywords.
- Thanks for the suggestion. We have modified the description of the reference [17] and [24]. (Please see the content marked in red for the description in reference [17] and reference [24] in Introduction Section)
- The figures are redrawn for 600 dpi. The new Figure 9 and Figure 11 are generated by using the program and has 600 dpi resolution. However, the letters in the Figures are still small.
- Thanks for the suggestion. The description of Object tensor has been modified in manuscript in new manuscript. (Please see the content marked in red for the description in Section 2.4).
- Thanks for the suggestion. Figure 8 and Figure 10 have been modified using Microsoft Easy Code Formatter.
- Thanks for the suggestion. We have changed the link to “Figshare” website and provide a DOI link.
Author Response File: Author Response.pdf
Reviewer 3 Report
A major linguistic revision is required.
There are many lexical errors. For instance, I have found seven occurrences of 'paring tree', rather than 'parsing tree'.
There are also several sentences whose syntax makes the text difficult to understand. For instance, this sentence in the abstract can be improved: 'However, there are lots of problems occurs such as pretending learning but actual absence and the behavior of plagiarism'.
The choice of terms is in several sentences difficult to understand. In the following sentence contained in the introduction 'To improve the logical concept of young people ..' the authors perhaps mean 'To improve the logical reasoning capabilities of young people ...'.
Experiment's protocol
I have not found a comprehensive description of the experiment's protocol. The authors mention that 206 students were involved in two experiments but I have not found a clear description of the duration of each test, how many tests were performed, how they were temporally distributed, how many students took part in each test, what was the objective of each test and so on.
Figures and equations
I was unable to understand the relationship between equation (1) and Figure 4. If t_ci is a function of t_si I would expect to have t_si on the abscissa and t_ci on the ordinate. On the contrary, we have t_ci on the abscissa and nothing on the ordinate. Where is t_max in the figure? Similar considerations apply to other figures.
Parsing trees and object tensors
I have not been able to understand the process through which a parsing tree is used to create an object tensor. This is a crucial step. I would recommend an improvement of the explanation and a table/figure that summarise the entire process
Plagiarism
According to the title and abstract the article is about predicting class loyalty and plagiarism, but the two experiments described in the Experimental Results section are about class loyalty, not plagiarism. Plagiarism is discussed in the Plagiarism Checking section but this is not enough to put plagiarism in the article's title.
Author Response
Response to Reviewer 3 Comments
There are many lexical errors. For instance, I have found seven occurrences of 'paring tree', rather than 'parsing tree'.
There are also several sentences whose syntax makes the text difficult to understand. For instance, this sentence in the abstract can be improved: 'However, there are lots of problems occurs such as pretending learning but actual absence and the behavior of plagiarism'.
The choice of terms is in several sentences difficult to understand. In the following sentence contained in the introduction 'To improve the logical concept of young people ..' the authors perhaps mean 'To improve the logical reasoning capabilities of young people ...'.
Response: Thanks for the useful comments. There are some answers to the questions.
The English of this paper is polished by native speaker.
Point 1: Experiment's protocol
I have not found a comprehensive description of the experiment's protocol. The authors mention that 206 students were involved in two experiments but I have not found a clear description of the duration of each test, how many tests were performed, how they were temporally distributed, how many students took part in each test, what was the objective of each test and so on.
Response 1: Thanks for the comment. The experimental part has been updated. (Please see the content marked in red for the description in Section 4.1)
Point 2: Figures and equations
I was unable to understand the relationship between equation (1) and Figure 4. If t_ci is a function of t_si I would expect to have t_si on the abscissa and t_ci on the ordinate. On the contrary, we have t_ci on the abscissa and nothing on the ordinate. Where is t_max in the figure? Similar considerations apply to other figures.
Response 2: Thanks for the comment. The figures have been updated.
Point 3: Parsing trees and object tensors
I have not been able to understand the process through which a parsing tree is used to create an object tensor. This is a crucial step. I would recommend an improvement of the explanation and a table/figure that summarise the entire process.
Response 3: Thanks for the comment. The descriptions of Object Tensor and descriptions of Parsing tree to OT are updated, and Table 1 is added. (Please see the content marked in red for the description in Section 2.4 and 3.2)
Point 4: Plagiarism
According to the title and abstract the article is about predicting class loyalty and plagiarism, but the two experiments described in the Experimental Results section are about class loyalty, not plagiarism. Plagiarism is discussed in the Plagiarism Checking section but this is not enough to put plagiarism in the article's title.
Response 4: Thanks for the comment. We add some description of the plagiarism checking part (Please see the content marked in red for the description in Section 4.2)
Author Response File: Author Response.pdf
Round 2
Reviewer 2 Report
Dear authors,
I have carefully read the changes you made to the proposed article.
I would advise you to pay attention to the divisions in syllables at the end of the lines.
Good luck in the future
Calin-Adrian COMES
Comments for author File: Comments.pdf
Author Response
Dear authors,
I have carefully read the changes you made to the proposed article.
I would advise you to pay attention to the divisions in syllables at the end of the lines.
Good luck in the future
Response: Thanks for the comment. Though the divisions in syllables at the end of the lines are done by the “Word” itself, we still check all of them one by one and don’t find anything weird. Good luck to you, too.
Reviewer 3 Report
- I do not know what a 'paring tree' is. The authors only replied 'The English of this paper is polished by native speaker.' Unfortunately, this does not help me to understand what a 'paring tree' is. Hence, I still guess it should be 'parsing tree'.
- In equation (1) when t_si=u_e the value of t_ci=1 (correct). But when t_si=u_max the value of t_ci is not zero but rather 1-(u_max-u_e)/t_max. Are the authors assuming that u_max-u_e=t_max?
Author Response
Response: Thanks for the useful comments. There are some answers to the questions.
Point 1: 1. I do not know what a 'paring tree' is. The authors only replied 'The English of this paper is polished by native speaker.' Unfortunately, this does not help me to understand what a 'paring tree' is. Hence, I still guess it should be 'parsing tree'.
Response 1: Thanks for the comment. That is parsing tree. I have fixed the typos.
Point 2: In equation (1) when t_si=u_e the value of t_ci=1 (correct). But when t_si=u_max the value of t_ci is not zero but rather 1-(u_max-u_e)/t_max. Are the authors assuming that u_max-u_e=t_max?
Response 2: Thanks for the comment. No, it is not zero but a small value. That is intentional. That design is used for observation. It makes a difference between the ones who do cheat when coding and the ones who just don’t want to learn. Though it will get the same grades of fuzzy results, we can still distinguish these two kinds of students by observing the values. For us, they all need us to give a further help, but the treatments to these two kinds of students are different.
Author Response File: Author Response.pdf