A Secure Data Publishing and Access Service for Sensitive Data from Living Labs: Enabling Collaboration with External Researchers via Shareable Data
, Evdokimos Konstantinidis 3,4,†, Gorka Epelde 2,5, Francisco Londoño 2, Despoina Petsani 3, Michalis Timoleon 3, Vasiliki Fiska 6, Lampros Mpaltadoros 6, Christoniki Maga-Nteve 6, Ilias Machairas 3 and Panagiotis D. Bamidis 3
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsAn overall interesting paper that deals with challenges of privacy in research, and proposes a solution in the form of their platform. The platform is well documented, and all its components are very well explained, making it very useful for any future implementations by research institutions that deal with similar privacy problems.
These are some of my recommendations:
The abstract is a bit too long, its information needs to be more condensed. Likewise, it would be preferable if the subject and goal of the paper were more clear from the abstract itself.
Introduction contains way too many abbreviations, consider using the full names of things, at least in the introduction part of the paper.
If possible, maybe the examples for anonymized data and synthetic data could use the same dataset, it would make the whole thing more compact and easier to understand. Naturally, this change is purely optional, and I would leave it at your discretion.
Likewise, since the focus in the paper has been on synthetic data generation with your platform(at least that was my impression), maybe the part of the experiment that deals with purely anonymized data is unnecessary.
You mention that you are using a blockchain for the VITALISE RAI server, maybe you could offer an explanation as to why you are using blockchain for this part of the system.
Author Response
Please see the attachment
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for AuthorsThe authors aim to facilitate research regarding Living Labs while preserving subjects’ privacy. Therefore, they introduce a service for secure data sharing of anonymized or synthetic versions of the original sensitive data to external researchers. They also propose remote experiment execution on the original data. This feature enables external researchers to design their own experiments and upload them for execution. Results are obtained through a cloud server. Finally, the authors present an adequate evaluation of the proposed system.
The paper is well-organized and well-written. It is technically sound, and it has merit.
Please check citation [1] in the Discussion.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 3 Report
Comments and Suggestions for AuthorsThe authors show in this article part of the work they have been developing in the European project VITALISE (https://vitalise-project.eu/).
It is a work that unites the concept of Living Labs + health, developing a working methodology based on the deployment of a solution developed by the authors that aims to ensure that the consumption of the data generated can be used in a secure, anonymous, and standardized way.
In addition, they show a basic statistical analysis that helps to make better decisions.
The results are shown on a proof-of-concept basis, with a controlled and fully characterized amount of data. The feeling is that it is scalable to real situations where data intake multiplies exponentially.
The reviewer considers that it is an interesting work for the scientific community, which shows how a controlled and secure management of health-related data can be carried out, whose main problem is the confidentiality of such data for the use of third parties.
According to the reviewer, this is a clear example of a term that is currently being advocated in Europe and that is Data Space. At no point throughout the paper is it cited, even if only subtly. Of course, in this case, the development is based on the Livings Labs and the data obtained from them. It is not intended to monetize the use of the data, but the structure presented is not far from a basic Data Space structure. Some components are missing, but it shows the standardization made to the structure of the generated data and the possibility of interoperability thanks to it.
On the other hand, although the authors show a basic statistical analysis, they should propose predictive analyses that allow decision-making optimization as they acquire more data.
It is interesting work, which must be implemented in a real way to be able to see the final potential it has.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
