Review Reports

Round 1

Reviewer 1 Report

Please find below a few comments on the article:

1. The authors were still referring to the study as "proposed" Why? (Line 37)

2. Change "agriculture" to "agricultural"  (Line 94)

3. What informed the sizes of your data population for each farmers' group? (Line 145-146)

4. Change "conducting" to "administering" (Line 157)

5. Remove "The Elobw Method Graph" on top of each of Fif 4 (a), (b) and (c) (Pages 7 & 8)

6. Recast Lines 356 -358 on Page 14. The sentences are incomplete

7. Recast Line 370 on Page 14. The sentence is incomplete

 

 

 

Author Response

Please see the attachment (word file)

Author Response File: Author Response.docx

Reviewer 2 Report

Strong Points:

1. The problem is important and well-motivated.

 2. In general the references are appropriate but can be improved.

 3. The method is innovative.

 4. The figures are appealing and have good visibility.

Weak Points:

 1. There are English issues. The writing style of this paper is not good enough. Authors should spend some time improving it so that the paper can be read more smoothly.

 2. The related work section must be enhanced. The related work section is not well organized. Authors must try to categorize the papers and present them in a logical way. The authors should explain clearly what  the differences are between the prior work and the solution presented in this paper.

 3. Experimental evaluation must be improved.

 4. The problem definition is not formal enough. The authors should add a clear and detailed problem definition. Authors should give a clear formal definition of the problem. The authors should add an example to illustrate the problem definition. 

 5. Some improvements are needed in the description of the method. For instance, K-means has trouble clustering data where clusters are of varying sizes and density. To cluster such data, you need to generalize k-means. Clustering outliers, Centroids can be dragged by outliers, or outliers might get their own cluster instead of being ignored. In addition, The authors should first give an overview of their solution before explaining the details.

6. It is beneficial to explain why K-means is used, instead of  KNN (k-nearest neighbors), Hierarchal clustering, Anomaly detection, Neural Networks.

7. The paper has some typos. Authors need to proofread the paper to eliminate all of them.

8. Some sentences are too long. Generally, it is better to write short sentences with one idea per sentence.

9. The algorithm of K-means should be included using pseudocode.

Author Response

Please see the attachment (word file)

Author Response File: Author Response.docx

Reviewer 3 Report

Some of the concerns expressed to the authors are addressed. Some are still ongoing. 

In Figure 2, the texts are on the axes. Table headings are out of alignment. Using K-means clustering alone is not enough to compare the results obtained. It is recommended to at least be compared with a different clustering algorithm.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 3 Report

In my opinion the revised paper can be accepted for publication