Automatic Generation Mechanism of Cause-Effect Graph with Informal Requirement Specification Based on the Korean Language

Round 1

Reviewer 1 Report

The authors propose an approach to transform and analyse informal system requirements of users or customers specified in Korean in order to automatically generate test cases from those. What makes it more complicated is that there is little research done for the Korean language-based requirement specifications. Developing such a system is a signifant challenge and would provide benefits for multiple fields. The solution proposed by the authors combines automatic simplification of informal requirements, generation of C3 Tree model and cause-effect generation.

The paper is very technical with most of the focus being on providing full step-by-step modelling. Unfortunately, major adjustments should be considered to improve the quality of the paper before it could be published:

1) The paper is very difficult to follow with many short sentences breaking up the flow. For example, abstract should be rephrased and double-checked.

Line 14: “In Korea, what is worse, there are few researchers for 14 automatically reducing requirement complexity”. I assume that sentence means that not much research is being done in the field but it says “researchers”.

Line 16: “Why need requirement simplification?” Check grammar here, more on that later.

Lines 17-19: “To solve this, we propose…” and “To do this, we propose…” Authors are using same words and structure, needs rephrasing.

2) Continuing on the previous point, this paper needs to be double-checked with regards to the correct use of English language. There are many grammar mistakes that need to be sorted. For example:

Line 30: “it is very difficult the automatic generation of test case”. Should be “difficult to automatically generate”?

Line 47: “Chapter 4 mentions our Korea Requirements”. Should be “Korean”?

Line 53: “The merits of the Cause-Effect Graph can design”. The “merits can design” or maybe the actual graph can help with designing something?

Line 78: “as Figure 3”. Should be “As shown in Figure 3”?

Line 139: “any front position of a sentence dislike English sentence”. Should be “unlike” instead of “dislike”?

Line 223: “we show the whole steps”. Should be “show all the steps”?

Overall, there are many mistakes throughout the whole text and extensive editing of English language and style is required.

3) Related studies section is very brief with almost no different methods mentioned. It contains many references to the mentioned Gary E. Mogyorodi works and also self-reference to “Sujeong Woo, Hyunseung Son, Wooyeol Kim, Jaeseung Kim, R. Youngchul Kim. “A Study Testcase 285 Extraction based M&S for Pre-Testing”. Korea Conference on Software Engineering 2012;14(1):181-183.”

Is there anything else out there? Again, the latest reference in this paper dates to 2014. Either more research needs to be done, or it should be explicitly mentioned that there is no other research in the field and explained why.

4) Section 3. How did you come up with the methods that are used in the proposed approach? No explanations or comparisons with other methods provided. Need to explain why specific techniques are used on each step of the proposed method. For example:

Line 131 – why is that specific corpus normalisation used?

Lines 216-220: why are those specific languages and tools are used? There is a brief mention that “Mecab-ko” is used for high speed-to-accuracy metric, but there is nothing about other tools.

5) The paper looks like a technical report with the method applied to just one very specific use case. How do you measure the system efficiency? How do you know that the results produced are correct? Is there a comparison between acquired model and something similar in the field? All that needs to be provided to show the actual benefit and workability of the developed method. Currently, paper has almost nothing to prove that the method actually works. This is the most important adjustment that needs to be made before the paper can be considered for publishing.

6) Some discussion and exact examples on where the developed method can be utilised should be provided. That will improve the quality of the conclusion and add value to the paper.

7) All reference are very old, some are pre-2000 and the latest is 2014. If there is nothing more recent on the problems described, then it should be specifically mentioned in the Related Studies section and explained why it is the case.

8) References are used incorrectly many times.

Lines 37 and 63: “Gary [3,4] mentions that the” and “Our previous research did follow Gary’s approach”. Gary is a first name of the author – Gary E. Mogyorodi, therefore his last name should be used in this case.

Line 51 says: “To make the Cause-Effect Graph, they did not clearly mention”. Who are “they”? Is that about the same paper by Gary E. Mogyorodi? Then he is the only author of his paper, and B.Math is his academic degree, not another person’s name.

Overall, references need to be double-checked and appropriate referencing style should be used.

 

Once these major issues are amended, the paper will be suitable for consideration for publication in this journal.

Author Response

Reviewer 1

 

The authors propose an approach to transform and analyse informal system requirements of users or customers specified in Korean in order to automatically generate test cases from those. What makes it more complicated is that there is little research done for the Korean language-based requirement specifications. Developing such a system is a signifant challenge and would provide benefits for multiple fields. The solution proposed by the authors combines automatic simplification of informal requirements, generation of C3 Tree model and cause-effect generation.

The paper is very technical with most of the focus being on providing full step-by-step modelling. Unfortunately, major adjustments should be considered to improve the quality of the paper before it could be published:

1) The paper is very difficult to follow with many short sentences breaking up the flow. For example, abstract should be rephrased and double-checked.

Answer) We did use “Grammarly.com” to fix our paper. After that, the English editor also did double-check this paper.

Line 14: “In Korea, what is worse, there are few researchers for 14 automatically reducing requirement complexity”. I assume that sentence means that not much research is being done in the field but it says “researchers”.

Answer) We changed the sentence based on your opinion.

“There are few kinds of research in Korea for automatically reducing requirement complexity”

Line 16: “Why need requirement simplification?” Check grammar here, more on that later.

Answer) We changed the sentence based on your opinion.

“Why do we need requirement simplification? Requirement complexity causes analyzers less readability and low understandability.”

Lines 17-19: “To solve this, we propose…” and “To do this, we propose…” Authors are using same words and structure, needs rephrasing.

Answer) We changed the sentence based on your opinion.

“To do this, we propose…”

2) Continuing on the previous point, this paper needs to be double-checked with regards to the correct use of English language. There are many grammar mistakes that need to be sorted. For example:

Line 30: “it is very difficult the automatic generation of test case”. Should be “difficult to automatically generate”?

Answer) We changed the sentence based on your opinion.

“it is difficult to automatically generate test cases”

Line 47: “Chapter 4 mentions our Korea Requirements”. Should be “Korean”?

Answer) We changed the sentence based on your opinion.

“Chapter 4 mentions our automatic generation mechanism of the Korean Requirements”

Line 53: “The merits of the Cause-Effect Graph can design”. The “merits can design” or maybe the actual graph can help with designing something?

Answer) We changed the sentence based on your opinion.

“The merits can design “

Line 78: “as Figure 3”. Should be “As shown in Figure 3”?

Answer) We changed the sentence based on your opinion.

“as shown in Figure 3”

Line 139: “any front position of a sentence dislike English sentence”. Should be “unlike” instead of “dislike”?

Answer) We changed the sentence based on your opinion.

“unlike English sentences”

Line 223: “we show the whole steps”. Should be “show all the steps”?

Answer) We have changed the contents of Chapter 4 from "Our Korean Requirement Analyzer for Cause-Effect Graph (KRA-CE)" to "Automatic generation mechanism".

Overall, there are many mistakes throughout the whole text and extensive editing of English language and style is required.

Answer) We changed the sentence.

3) Related studies section is very brief with almost no different methods mentioned. It contains many references to the mentioned Gary E. Mogyorodi works and also self-reference to “Sujeong Woo, Hyunseung Son, Wooyeol Kim, Jaeseung Kim, R. Youngchul Kim. “A Study Testcase 285 Extraction based M&S for Pre-Testing”. Korea Conference on Software Engineering 2012;14(1):181-183.”

Is there anything else out there? Again, the latest reference in this paper dates to 2014. Either more research needs to be done, or it should be explicitly mentioned that there is no other research in the field and explained why.

Answer) We added the sentence based on your opinion.

1: “Berk Bekirolu focused on testing with the cause-effect graph from English software specifications. Nobody does work with automatically generating Cause-Effect Graphs from the Korean requirements.”

2: We did search our research title with IEEE Xplore and ACM Digital Library. But there are no latest related papers.

`

4) Section 3. How did you come up with the methods that are used in the proposed approach? No explanations or comparisons with other methods provided. Need to explain why specific techniques are used on each step of the proposed method. For example:

Line 131 – why is that specific corpus normalisation used?

Answer) We added the sentence based on your opinion.

“It is challenging to analyze the meaning of informal sentences. In the case of Korean, it is difficult to identify the subject and object in the passive sentences. That is why we convert passive sentences into active sentences.”

Lines 216-220: why are those specific languages and tools are used? There is a brief mention that “Mecab-ko” is used for high speed-to-accuracy metric, but there is nothing about other tools.

Answer) We have changed the contents of Chapter 4 from "Our Korean Requirement Analyzer for Cause-Effect Graph (KRA-CE)" to "Automatic generation mechanism".

5) The paper looks like a technical report with the method applied to just one very specific use case. How do you measure the system efficiency? How do you know that the results produced are correct? Is there a comparison between acquired model and something similar in the field? All that needs to be provided to show the actual benefit and workability of the developed method. Currently, paper has almost nothing to prove that the method actually works. This is the most important adjustment that needs to be made before the paper can be considered for publishing.

Answer) We have changed the contents of Chapter 4 from "Our Korean Requirement Analyzer for Cause-Effect Graph (KRA-CE)" to "Automatic generation mechanism". In the next paper, we collect the right requirements samples, improve tools, and verify the developed tools.

6) Some discussion and exact examples on where the developed method can be utilised should be provided. That will improve the quality of the conclusion and add value to the paper.

Answer) we did totally change the whole picture of our automatic generation mechanism of the Korean Requirements Analyzer for Cause-Effect Graph in Figure 10.

7) All reference are very old, some are pre-2000 and the latest is 2014. If there is nothing more recent on the problems described, then it should be specifically mentioned in the Related Studies section and explained why it is the case.

Answer) We added the sentence based on your opinion.

“Berk Bekirolu focused on testing with the cause-effect graph from English software specifications. Nobody does work with automatically generating Cause-Effect Graphs from the Korean requirements.”

8) References are used incorrectly many times.

Lines 37 and 63: “Gary [3,4] mentions that the” and “Our previous research did follow Gary’s approach”. Gary is a first name of the author – Gary E. Mogyorodi, therefore his last name should be used in this case.

Answer) We fixed it.

Line 51 says: “To make the Cause-Effect Graph, they did not clearly mention”. Who are “they”? Is that about the same paper by Gary E. Mogyorodi? Then he is the only author of his paper, and B.Math is his academic degree, not another person’s name.

Answer) We fixed it.

Overall, references need to be double-checked and appropriate referencing style should be used.

Answer) We fixed it.

Once these major issues are amended, the paper will be suitable for consideration for publication in this journal. 

 

Author Response File: Author Response.docx

Reviewer 2 Report

The authors in the paper presented a complete method to automatically generate test cases with informal requirement documents in the Korean language.

The developed method consists of six steps, namely: simplify informal requirements, extract Cause and Effect from them, create C3Tree with the extracted Cause and Effects, generate the cause-effect graph, convert this graph into decision tables, and generate test cases based on the decision tables.

The presented workflow is exciting and no doubt gives good results.

However, I have a few comments and observations.

The literature review in the paper is poor. There should definitely be a more extensive description of other papers that deal with this topic.

A description of Figure 2 would be very valuable.

There is a lack of critical review of the method developed. It would be valuable to perform some tests that will give numerical results that show whether the strategy's result is error-free and satisfactory every time, even for complex cases.

Author Response

Reviewer 2

 

The authors in the paper presented a complete method to automatically generate test cases with informal requirement documents in the Korean language.

The developed method consists of six steps, namely: simplify informal requirements, extract Cause and Effect from them, create C3Tree with the extracted Cause and Effects, generate the cause-effect graph, convert this graph into decision tables, and generate test cases based on the decision tables.

The presented workflow is exciting and no doubt gives good results.

However, I have a few comments and observations.

The literature review in the paper is poor. There should definitely be a more extensive description of other papers that deal with this topic.

Answer) We added the sentence based on your opinion.

1: “Berk Bekirolu focused on testing with the cause-effect graph from English software specifications. Nobody does work with automatically generating Cause-Effect Graphs from the Korean requirements.”

2: We did search our research title with IEEE Xplore and ACM Digital Library. But there are no latest related papers.

`

 

A description of Figure 2 would be very valuable.

Answer) We added an exact example for Figure 2.

“Each meta model stores meta information about the model (XMI File). The Cause-Effect Graph is automatically converted to the Decision Table by Model Transformation Engine 1 with Transformation Rule 1. Model Transformation Engine1 automatically converts the Cause-Effect Graph generated by referring to the Cause-Effect Meta model to the Decision Table created by the Decision Table Meta model. Likewise, the Decision Table is automatically converted into a Test case by Model Transformation Engine 2.”

There is a lack of critical review of the method developed. It would be valuable to perform some tests that will give numerical results that show whether the strategy's result is error-free and satisfactory every time, even for complex cases.

Answer) We have changed the contents of Chapter 4 from "Our Korean Requirement Analyzer for Cause-Effect Graph (KRA-CE)" to "Automatic generation mechanism". In the next paper, we will collect requirements samples, improve tools, and verify the developed tools.

 

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

Dear Authors,

 

Thanks for revising your paper. It is appropriate for publication now. However, some feedback was not fully addressed and further minor changes are needed as follows:

 

1) Line 17 - 19. You are using same sentence structure "To do this, we propose.." in 2 consecutive sentences. That needs rephrasing, do not use same wording in the nearby sentences, that breaks the flow.

2) Line 55. You still write "merits can design". Merits cannot design anything, people can utilise the benefits of certain methods and design something based on that. That is improper use of English language. Please change that sentence accordingly.

3) In the previous review I mentioned the following: "Some discussion and exact examples on where the developed method can be utilised should be provided. That will improve the quality of the conclusion and add value to the paper."

Unfortunately, in the conclusion or introduction there are still no examples of where you can use the proposed methodology. Therefore, introduction and conclusion need to be expanded to show the exact application examples of this method.

 

Once those minor changes are amended, the paper will be suitable for publication in the journal.

 

 

Author Response

1) Line 17 - 19. You are using same sentence structure "To do this, we propose.." in 2 consecutive sentences. That needs rephrasing, do not use same wording in the nearby sentences, that breaks the flow.

Answer) We changed the sentence based on your opinion.

Before sentence: “To do this, we propose the following procedures:”

Changed sentence: “which works the following procedures:”

2) Line 55. You still write "merits can design". Merits cannot design anything, people can utilise the benefits of certain methods and design something based on that. That is improper use of English language. Please change that sentence accordingly.

Answer) We changed the sentence based on your opinion.

Before sentence: “The merits can design the minimal test case…”

Changed sentence: "With Cause-Effect Graph, we can create the minimal test case…"

3) In the previous review I mentioned the following: "Some discussion and exact examples on where the developed method can be utilised should be provided. That will improve the quality of the conclusion and add value to the paper."

Unfortunately, in the conclusion or introduction there are still no examples of where you can use the proposed methodology. Therefore, introduction and conclusion need to be expanded to show the exact application examples of this method.

Answer) We add a case study because we explain our mechanism with actual requirements of ‘A guidance for completing DPS for a Missile system’.

Thank you for your considerate review.

Author Response File: Author Response.docx