Development of a Model Using Data Mining Technique to Test, Predict and Obtain Knowledge from the Academics Results of Information Technology Students

Round 1

Reviewer 1 Report

The authors certainly improved the text and corrected it according to most of my requests. Since this is a resubmission, it should be corrected according to the requirements of other reviewers. Maybe I'm wrong but I don't notice that here.

Authors should definitely check the text and documents, typos

ie:

Line 64: envornment  = >  environment

146: algortihm  => algorithm    

167: accurracy => accuracy

Author Response

Reviewer 1

Comments and Suggestions for Authors

The authors certainly improved the text and corrected it according to most of my requests. Since this is a resubmission, it should be corrected according to the requirements of other reviewers. Maybe I'm wrong but I don't notice that here.

Authors should definitely check the text and documents, typos

ie:

Line 64: envornment  = >  environment

146: algortihm  => algorithm    

167: accurracy => accuracy

Response and revisions (line numbers):

In regards to the requirement of other reviewers, I have made the necessary correction and the changes has been effected in the newly edited manuscript submitted.

line 64: The error has been corrected

line 146: The spelling error has been corrected

line 167: The spelling error has been corrected

Author Response File: Author Response.docx

Reviewer 2 Report

Dear authors, I appreciate having taken into account my humble suggestions in order to improve your excellent work.

I have been able to review the new version of your work and I have been able to verify that you have made an effort to be able to consider the recommendations made.

Your work is now a more consistent , much clearer and very easy to understand for the reader.
In my opinion, I consider that your work is ready to be published.

Greetings

Author Response

Reviewer 2

Comments and Suggestions for Authors

Dear authors, I appreciate having taken into account my humble suggestions in order to improve your excellent work.

I have been able to review the new version of your work and I have been able to verify that you have made an effort to be able to consider the recommendations made.

Your work is now a more consistent , much clearer and very easy to understand for the reader.
In my opinion, I consider that your work is ready to be published.

Greetings

Response and revisions (line numbers):

Thanks so much for the comments. I am grateful for your humble suggestions.

Author Response File: Author Response.docx

Reviewer 3 Report

Thank you for addressing all my comments last time. 

1. In section 3.2, it mentions using measures including recall, precision, and F-measure to evaluate the model fitting. However, no results were presented using these three measure.
2. In my last comments, RMSE, RA, and RRSE shouldn't be used in Table 1 since it was for the continuous outcome variable. Should this table present results in recall, precision, F-measure instead?
3. For figures 8-12, it seems that a certain ethnic group of students are only from one university, e.g., Nigerian students are all from Lagos State University Nigeria. Then, why are we seeing five curves in Figure 8? Supposed that each curve stands for one university. 
4. Also for Figures 8-12, please use more informative labels rather than A, B, C, D, and F to annotate each university. 

Author Response

Reviewer 3

Comments and Suggestions for Authors

Thank you for addressing all my comments last time. 

1. In section 3.2, it mentions using measures including recall, precision, and F-measure to evaluate the model fitting. However, no results were presented using these three measure.
2. In my last comments, RMSE, RA, and RRSE shouldn't be used in Table 1 since it was for the continuous outcome variable. Should this table present results in recall, precision, F-measure instead?
3. For figures 8-12, it seems that a certain ethnic group of students are only from one university, e.g., Nigerian students are all from Lagos State University Nigeria. Then, why are we seeing five curves in Figure 8? Supposed that each curve stands for one university. 
4. Also for Figures 8-12, please use more informative labels rather than A, B, C, D, and F to annotate each university. 

Response and revisions (line numbers):

Thanks so much for the comments. I am grateful for your humble suggestions. The table 1 has been changed with the values of the recall, precision and F- measure.   For figure 8-12, shows the plot of the TP, FP, Recall, Precision, F-measure, MCC Area obtained for the Information Technology (I.T) students in Lagos state University Nigeria; University of Kirkuk, Iraq; University of Cape Town, South Africa; Sudan University of science and technology; and Jawaherlal Nehru University, India. The RMSE, RA and RRSE have been removed from the table 1 as we agreed that this is for continous outcome variable based on your suggestions. The labels has been corrected to indicate that universities as listed above.  

Author Response File: Author Response.docx

Round 2

Reviewer 3 Report

Thank you for addressing my comments. I don't have any further comments. 

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

In this paper Authors used WEKA to analyze the students’ academic performance of Information Technology department in five universities across five countries which includes Iraq, Sudan, Nigeria South Africa and India. Some conclusions are given. The paper could be interesting but the text cannot be accepted in the present form.

General remark: If the authors are already using the available WEKA environment and there is nothing new there, the authors should expand the analysis of the results and discussion in order to give something new from the text. Technical shortcomings also need to be corrected.

line 59: authors should cite WEKA. Something like "... Waikato Environment for Knowledge Analysis (WEKA) [REfs.] software to analyze ..."

[Ref] S.R. Weka Garner, The waikato environment for knowledge analysis, in: Proceedings of the New Zealand Computer Science Research Students Conference, 1995.

[Ref] E. Frank, et al., Weka-a machine learning workbench for data mining, in: Data Mining and Knowledge Discovery Handbook, Springer, 2009, pp. 1269–1277

line 66: Ref. [9] => [8]

line 67: Ref. [8] => [9]

line 94: Ref [18] appears before Ref [11] ?

Subsubsections 3.1.1-3.1.5 should be better designed or moved in appendix.

Currently figures are unacceptable and need to be fixed. More precisely they should be enlarged and/or text on their axes.  Only at a zoom of 200% something is visible.

Figure 4 should be  Figure 5, ...etc. Where is figure 3 and its caption. I cant find.

General shortcoming: Figures need to be described better and more extensively. This will raise the discussion to a higher level.

Tables 1-5 should somehow be merged into one. That would make it easier to see and conclude something

Reviewer 2 Report

Dear authors, the theme of the work is very interesting and will definitely have a valuable contribution in the field of Data Mining and Machine Learning applied to education (EDM) especially applied to the quality of higher university service.
However, I am sending you some of the observations made as a result of the detailed reading of your research work:
Regarding the summary, in lines 16 and 17 you state that one of your objectives is to discover patterns of knowledge, however in your results and discussions you do not mention that.
In the introduction they do not refer to any work focused on the J48 algorithm or its applications in education, it would be very convenient if they did.
In the data description section: They do not make a clear description of the data, they could use a figure or a table with the metadata, in order to better understand the data.
In Figure 2, I find the file format details unnecessary, rather they should focus more on the full path of data analysis. I think that they unnecessarily focus the work around the WEKA tool and do not try very hard to explain and detail the nature of the data and the algorithm used, I suggest that they focus more on the scientific aspect of the method and algorithm used. Nor do I find a clear explanation of the route followed in the research work and the alignment to the KDD.
In section 3 (Methods), it would be interesting if you could experiment with: larger datasets, with other classification methods and evaluate the differences between them. This would add a very interesting comparative analytical approach to the work. I think it would not be necessary to show all the Weka reports, using a summary table would be the most appropriate.
In section 3.3. Again, I consider that it is not necessary to show all the Weka reports regarding the results of the five universities. Only a table or two with the most significant values ​​would be the most recommended.
In the results and discussion section I do not find a deep and detailed analysis of the results, their indicators and evaluation metrics, much less an interpretation applied to the case study. I strongly suggest that you do.
I recommend you can improve those aspects, to give more value to the interesting work you submitted.
Greetings

Reviewer 3 Report

1. Why were data from each school analyzed separately rather than combining them by adjusting students' ethnicity group (location of schools)?
2. In Section 3, it is not clear what variables were used in the model, and how the splitting of data into training and test was done. 
3. It seems that the outcome variable is whether the students pass the exam or not. In that case, RMSE, RA, and RRSE shouldn't be used since these metrics were used for predicting continuous variables instead of binary variables. 
4. In section 3.1, instead of showing the screenshots of the results, it is recommended to summarize them into tables or figures. Screenshots can be included in supplementary files.
5. In section 3.2, since terminologies are commonly used in the field, it is unnecessary to explain them here. 
6. In section 4, all tables from Table 1 to Table 5 can be merged into one table since they share the same columns. Please do so and add another column to indicate the ethnicity of students, and the number of samples in training and test datasets. 
7. Meanwhile, please remove Figures 14 to 18 since it provides the redundant information as Table 1 to Table 5.