Corporate Data Sharing, Leakage, and Supervision Mechanism Research

Round 1

Reviewer 1 Report

This study aims to develop a data supervision model that considers data sharing, data leakage, and derived profits and losses for stable data sharing environments. The biggest contribution of this paper is that these paper summarized factors that companies should consider when they decide to share data with other companies. However, it is expected that there is a better study if you reflected the following modifications below.

 

1. In abstract section, the content of the research methodology is too long. It would be better to reduce the content of the research methodology and write more details about the results of the study.
2. In Section 2 Theoretical Background contains too much contents. I recommend dividing into two sections to improve reader's understanding.
3. You should remove the "(  )" in the whole text. e.g. line 104, line 123.
4. Please, correct all typo errors e.g. line 137. All referencing in the manuscript should have a space before "[" character.
5. It is hard to know what the Supervision model means. If the word is originally used, reference should be added. And if the word newly used in this paper, you should mention some descriptions or definitions of the word.
6. You should add the reference the contents of the Harvard Business Review used for validation in Section 5.1.
7. In order to validate the proposed formula, the example of Harvard Business Review was applied. But, you should clearly write that you referenced value. e.g. amount of leaking data. In addition, you should write the application of this research should be written in detail. When the formula proposed in this paper applied to the Harvard Business Review, there is no exact numbers and the process of this validation is not convincing. If possible, you should write them in a separate section.

Author Response

Response to Reviewer 1 Comments

 

 

 

Point 1: In abstract section, the content of the research methodology is too long. It would be better to reduce the content of the research methodology and write more details about the results of the study.


 

Response 1: Thank you so much for your valuable viewpoint, and we have totally modified the abstract. Please see the abstract section of the manuscript.

Data sharing helps to make full use of each other's data and enable the digital economy. With the gradual rise of corporate data sharing, the frequent occurrence of data leakage events highlights the dilemma of data sharing, leakage and supervision, which infringes on the data assets of the leaked party. Our paper aims to design an effective data supervision mechanism and achieve the stability of data sharing alliance. Therefore, this paper considers the data technology capabilities of both parties of the data sharing alliance and the benefits and loss of data leakage, establishes the game model and supervision mechanism of data sharing, leakage and supervision under the condition of complete information, analyzes the game equilibrium and the influence of different supervision levels on the company's data sharing and leakage behavior. The results show that the company sharing and leaking behavior is affected by both the level of data supervision fines and the relative technical level. Our model can make up for the weakness of the low-tech company, control the company's choice of leaking behaviors, and ensure the stability of data sharing alliance by designing reasonable data supervision mechanism, especially the severe data supervision strategy.

 

Point 2: In Section 2 Theoretical Background contains too much contents. I recommend dividing into two sections to improve reader's understanding.

 

Response 2: Thank you so much for your suggestion. We have re-organized the paper as follows:

We review the previous relevant literatures in section 2. Because the original theoretical background is too long to understand, this part is reorganized according to the content and divided into three sub-sections: 2.1.data sharing, 2.2.data leakage and 2.3.data supervision. In the corresponding part, the corresponding theory research is introduced. Please see from page 3, line 98 and 116, to page 4, line 149.

 

Comment 3: You should remove the "(  )" in the whole text. e.g. line 104, line 123.

 

Response 3: Thank you so much for your viewpoints, and we agree with your great ideas.

We have removed the "(  )" in the whole text. Please see page 2, line 70, page 3, line 119 and 138, page 7, line 241, page 12, line 400, page 13, line 417.

 

Comment 4: Please, correct all typo errors e.g. line 137. All referencing in the manuscript should have a space before "[" character.

 

Response 4: Thank you so much for your careful work.

We have corrected all the typos. Please see the new word e.g. line 153, line 185, line 246, ref [6], ref [43]. And we increase the space before "[" character.

 

Comment 5: It is hard to know what the Supervision model means. If the word is originally used, reference should be added. And if the word newly used in this paper, you should mention some descriptions or definitions of the word.

 

Response 5: Thank you so much for your valuable comments.

As you suggested, we re-organized our paper, and add the descriptions of “Supervision model”, please see line 163 and the description as follow: In view of the current situation of data leakage and lack of supervision in company data sharing, this paper establishes a model of data sharing, leakage and supervision. The model is that before data sharing, both parties formulate a regulatory mechanism with fines as the main constraint, and supervise each other's corporate disclosure behavior, and use the regulatory mechanism to influence the company's behavior choice and solve the current problems.

 

Comment 6: You should add the reference the contents of the Harvard Business Review used for validation in Section 5.1.

 

Response 6: Thank you so much for your careful work.

We have increased the reference of the Harvard Business Review. In addition, we reorganized the case background section. Please see line 445.

 

Comment 7: In order to validate the proposed formula, the example of Harvard Business Review was applied. But, you should clearly write that you referenced value. e.g. amount of leaking data. In addition, you should write the application of this research should be written in detail. When the formula proposed in this paper applied to the Harvard Business Review, there is no exact numbers and the process of this validation is not convincing. If possible, you should write them in a separate section.

 

Response 7: This is a very valuable suggestion, thank you so much.

(1)As you suggested, we re-organized our paper. In order to express the reference values of simulation analysis data more clearly, all simulation data are listed in Table 3, Please see page 11. 

(2)Our simulation analysis part is to verify the effectiveness of the model and further discuss the impact of supervision mechanism on the behaviour of stakeholders, which is not directly related to the case, so the data of simulation part is not from the real data of the case. We realize that it is a dilemma as you argued, and therefore we put the case in the simulation analysis part to explain the problem. According to your suggest, we reorganize the structure of the article, put the case in the extension discussion part after the simulation analysis results, and give some management enlightenment from the combination of simulation results and cases. Please see page 11, 5 simulation analysis, and 5.2 Extension discussion of “SF Express Company: Data War". The probability for a given country h to be in a class k should be the proportion of observations (households) in country h that belong to the income class k. On page 9, the first equation (it would be easier for the reader if the equation is numbered) is not exactly the proportion of people because the authors take the sum of the probability. The interpretation of the equation in not obvious. Normally, after estimating a mixture of regression model we have for each observation its estimated probabilities to be classified into the different classes identified. What is often done is to classify a given observation into the class where its estimated probability is higher. In many software this is also the method used that gives us the proportion of people in each of the classes. The authors should explain the equation on page 9 and how to interpret it. Alternatively, they may use the proportion approach which will make the interpretation easier.

Author Response File: Author Response.docx

Reviewer 2 Report

1. Some writing errors and small typos were found in the manuscript, please correct and check all the probable errors and mistakes. (for example ref [35] , line 28 correct form is [2].)
2. Please give the disadvantages of the proposed techniques in the paper and how to overcome these disadvantages.
3. Future research directions should be provided in Conclusion Section.
4. Poorly formulated research questions.
5. All equations should be followed with symbol explanations and correct form of subscript
6. For all  symbol and parameters clearly mentions the units
7. References have to write in in correct form (for example ref [35])
8. That’s better use and add the references that recently published

 

Author Response

Response to Reviewer 2 Comments

 

 

 

Point 1: Some writing errors and small typos were found in the manuscript, please correct and check all the probable errors and mistakes. (For example ref [35], line 28 correct form is [2].)

Response 1: Thank you so much for your careful work.

We have corrected all the typos. Please see the new word (for example line 35 correct form is [2], ref [35], line 153, line 185, line 246.)

Point 2: Please give the disadvantages of the proposed techniques in the paper and how to overcome these disadvantages.

Response 2: Thank you so much for your viewpoints, and we agree with your great ideas.

We have increased the techniques in the paper and how to overcome these disadvantages in future research. Please see line 524 and the description as follow:

First of all, on the technical level, the disadvantage of this article is to use the theory of complete information game to build a supervision model, but considering that it is difficult to know all the information of each other in the actual company data sharing. In the next study, we will consider the data sharing, leakage, and supervision in non-complete information situations.

Point 3: Future research directions should be provided in Conclusion Section.

Response 3: Thank you so much for your valuable comments.

As you suggested, we re-organized our paper. We add the future research directions at the 6.4 Future Research section, please see page 15 line 528. The description as follow:

In the future research direction, the model proposed in this paper requires both parties to supervise each other, which is also the reason for the imbalance of supervision in the conclusion. We also find that most enterprises' data sharing is based on the third-party data platform. Therefore, we can consider the further study on data supervision mechanism based on the third-party data platform in the future research.

Point 4: Poorly formulated research questions.

Response 4: Thank you so much for your suggestion; we have re-organized the paper as follows:

(1)We have totally modified the abstract. Please see the abstract section of the manuscript.

(2)We add the descriptions of “Supervision model”, please see line 163, and the description as follow: In view of the current situation of data leakage and lack of supervision in company data sharing, this paper establishes a model of data sharing, disclosure and supervision. The model is that before data sharing, both parties formulate a regulatory mechanism with fines as the main constraint, and supervise each other's corporate disclosure behavior, and use the regulatory mechanism to influence the company's behavior choice and solve the current problems.

(3) We realize that this is because we put the case in the simulation analysis part to explain the problem, so according to your opinion, we reorganize the structure of the article, put the case in the extension discussion part after the simulation analysis results, and give some management enlightenment from the combination of simulation results and cases. Please see page 11, 5 simulation analysis, and 5.2   Extension discussion of “SF Express Company: Data War".

Point 5: All equations should be followed with symbol explanations and correct form of subscript.

Response 5: Thank you so much for your careful work.

All the symbols used in this paper are given in Table 1, and all equations and the application of symbols in this paper are also explained in Chapter 3.2. Model analysis. And we corrected the form of subscript; please see page 7 line 232.

Point 6: For all symbol and parameters clearly mention the units.

Response 6: Thank you so much for your viewpoints.

The definition and value range of all symbols in this paper are described in Table 1. Because the symbols in this paper are data related, and the specific unit of the symbol also depends on different situations, the symbol in this paper does not set a unit. In the simulation analysis part, the specific value of the symbol has been modified to a more clear form, and the structure of the simulation part has also been adjusted. Please see page 10 Table 3.

Point 7: References have to write in in correct form (for example ref [35]).

Response 7: Thank you so much for your careful work.

We have corrected all the references typos. Please see page 16 to 18, ref [6], ref [35], and ref [43].

Point 8: That’s better use and add the references that recently published.

Response 8: Thank you so much for your valuable comments.

We have updated some references that published recently. Please see ref [2], ref [8], ref [9], and ref [13].

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

1. The Figure 1 must be high resolution.
2. The Table 3 should be formatted. They do not have the caption of table.
3. Hard to know about the validation or verification of your proposed model. Where is the verification section? I think paper only have proposed and discussion mentioned. The authors should be described verification process and results.

Author Response

Comment 1: The Figure 1 must be high resolution.

 

Response 1: Thank you so much for your valuable viewpoint, and we have replaced the Figure 1 to a high resolution. Please see page 4.

We referred MDPI for image format requirements: According to the Instructions for Authors of Sustainability, File for Figures and Schemes must be provided during submission in a single zip archive and at a sufficiently high resolution (minimum 1000 pixels width/height, or a resolution of 300 dpi or higher). Common formats are accepted, however, TIFF, JPEG, EPS and PDF are preferred. Please refer to the website link as follows,

https://www.mdpi.com/journal/sustainability/instructions#figures.  

Therefore, we have updated all the figures as follows:

1. According to the requirements, we have modified the Figure 1 as a new TIFF with 600×600 dpi for meeting the standard above. 
2. In addition, we also replaced the other figures in the document as a new TIFF with 600×600 dpi. Please see page 4 Figure 1, page 10 Figure 2, page 11 Figure 3 and page 12 Figure 4. 
3. At the same time, we added a single zip file for all the figures.

Comment 2: The Table 3 should be formatted. They do not have the caption of table.

 

Response 2: Thank you so much for your valuable viewpoint, and we have totally added the caption of table 3 as “Table 3. Simulation parameters of data sharing alliance”. Please see page 10, line #350. In addition, we have adjusted the way the parameters in the table are written to correspond to the symbol definition section above.

Table 3. Simulation parameters of data sharing alliance 

Scale of shared data	DA=DB=1000
Data quality	qA=qB=0.8
Proportion of leaking data	x=y=1
Degree of punishment	severe penalty ω =2 ，mild penalty ω=0.5
Relevance degree between data Leaked loss and benefits of Data Leaking	k=0.5

 

Comment 3: Hard to know about the validation or verification of your proposed model. Where is the verification section? I think paper only have proposed and discussion mentioned. The authors should be described verification process and results.

 

Response 3: Thank you so much for your suggestion. After reviewing the content of the article, we find that there is indeed a lack of verification of the propositions proposed above. As you suggested, we have added the description of the validation of proposed proposition, please see the page 11 5.1.1 Supervision mechanism for mild penalties, and page 12 5.1.2 Supervision mechanism for severe penalties. 

Above, we propose three propositions based on the calculation results of the model. The simulation verification for each proposition is illustrated as follows:

1. Proposition 1: In the first case, S₁ becomes the Nash equilibrium strategy. S₁ strategy combination is the optimal sharing strategy of data alliance, and the supervision penalty level condition is , otherwise it cannot maintain the optimal sharing strategy. (Please see page 8 line #272)(2)Proposition 2:If the fine is satisfied  or , one of the two companies may leak the other company’s data, while the other party does not leak, and the leaking data behavior only occurs in the company with higher data technology level. (Please see page 9 line #294)(3)Proposition 3:If the penalty level  of the supervision mechanism is bigger than the equilibrium point 1 (fig. 2), the S₄ equilibrium does not exist; if the penalty level is less than the equilibrium point, the S₁ equilibrium does not exist. (Please see page 9 #line 317)Based on the above analysis, the simulation results under two different punishment levels verify the three propositions mentioned above, and also prove that the supervision model and supervision mechanism proposed in this paper can influence the behavior of data sharing alliance participants. From the comparison of the two punishment levels, it can be seen that the greater the punishment level, the lower the risk of data leakage; meanwhile, in the same equilibrium, the benefit of the party who leaks data increases with the decrease of punishment level, while that of the party whose data is leaked decreases. Therefore, strict supervision can better protect the rights and interests of the leaked party and minimize the loss of data leakage. (please see page 12 #line 405)
2. In the mild penalty, the S₁ Nash equilibrium does not exist and cannot control the behavior of companies A and B leaking data, if the penalty level is less than 1, the S₁ equilibrium does not exist, which verifies that in Proposition 3. When the supervision mechanism is severe punishment mechanism, it can be seen from Table 3 that . In Figure 4 that S₄ equilibrium does not exist at this time, which verifies that in Proposition 3 , S₄ equilibrium does not exist. (Please see page 11 #line 362)
3. It can be seen from Table 3 that under the mild punishment mechanism, the punishment coefficient .From the distribution of horizontal axis t and three kinds of equilibrium that when the value of t falls within the interval ,  is difficult to meet the condition or , and it is S₄ equilibrium; under the severe punishment mechanism, the punishment coefficient . From the distribution of horizontal axis t and three kinds of equilibrium that when the value of t falls within the interval ,  is difficult to meet the condition , and it is S₁ equilibrium. In other cases, when  satisfies [0,0.5], Company B has higher technical level, The data leakage occurred only in Company B, it is S₂ equilibrium; when  satisfies [2,3], Company A has higher technical level, The data leakage occurred only in Company A, it is S₃ equilibrium. S₂ equilibrium and S₃ equilibrium are one party discloses data and the other party does not, and the leaking data behavior only occurs in the company with higher data technology level. This further verifies Proposition 2 proposed above. (Please see page 11 and page 12)
4. According to the horizontal axis distribution, S₁ equilibrium only exists when the relative technical level  is between , meets the condition of , and there is no data leakage behavior, which is in line with the conclusion of proposition 1. (Please see page 11 line #365)

Author Response File: Author Response.pdf

Round 3

Reviewer 1 Report

You did really good job. Thank you for reflecting all of my comments.
Thank you for your valuable research. I hope to see the superior research in future.