Dual-Safety Knowledge Graph Completion for Process Industry
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsThe manuscript proposes a knowledge graph method to integrate information on security risks in industrial processes and a link prediction task on the knowledge graph to discover previously unseen knowledge regarding potential security risks. The problems being addressed are interesting. However, the contribution of the manuscript is limited.
The application of the method proposed by the authors in the context of finding potential security risks is not well explained in the manuscript. The authors can add, for example, use cases and scenarios for a better explanation.
Contributions should be highlighted in the introduction.
Experimental results need improvement.
Comments on the Quality of English LanguageMinor editing of English language required
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for AuthorsIt is an interesting paper in the current evolution of cybersecurity systems.
However, I think the authors should try to present the performance of the research through a simpler model, with some details of the processes, before jumping to a more complex model. For example, zooming on Figure 2.a (crack hydrogenation process KG) cannot be read as are used different characters than the international alphabet (i.e., A, B, C etc.). Through a simpler model it can also be illustrated what results are correct and which ones should be eliminated or avoided by the proposed analysis.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 3 Report
Comments and Suggestions for AuthorsThe authors propose knowledge graphs to improve security by including processes and not only the technical part. The paper is easy to follow and good written. However, I'm unsure about the contribution. The results of the evaluation do not really reveal anything extraordinary, although it might help in some cases.
The authors could improve the evaluation (easier dataset, applicability to the use case of processes) and add a discussion on whether they reach their goal and which limitations their approach has.
The conclusion shortly summarizes the work done and future work. If 2) and 4) is improved, the current conclusion would probably fit. Currently, I miss statements about applicability to their use case.
Comments on the Quality of English LanguageMinor items, such as missing spaces.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
Comments and Suggestions for AuthorsThe authors have made efforts to improve the manuscript. No further comments.
Comments on the Quality of English LanguageMinor editing of English language required
Reviewer 2 Report
Comments and Suggestions for AuthorsThe authors improved the paper through the modified paragraphs. The explanations given between lines 288 - 306 help the reader in understanding the depth of the analysis performed and the complexity of the models approached by the authors.
Reviewer 3 Report
Comments and Suggestions for AuthorsThank you for incorporating the comments.