Property-Based Semantic Similarity Criteria to Evaluate the Overlaps of Schemas

Round 1

Reviewer 1 Report

Dear authors,

TThis article addresses an interesting topic that could be of interest for the community. The article is well written, and worthwhile paper that deserves publication. However, the presentation of the problem is not addressed logically and without following an adequate order throughout the different sections. For example, the analysis of the state of the art is provided in the end of the paper when what is expected is to see it at the beginning to be able to start from a base point.

I am adding a few comments and possible recommendations to improve the article in each of the sections. 

Title

I am not sure if it reflects the essence of the article, which seems to be more focused on the semantic similarity of the schemas from the properties perspective. For example, something like: “A property based semantic similarity criteria to evaluate the overlaps of schemas”, it would be more related to the content.

Abstract

The abstract is appropriate and reflects what readers would expect to find in the article. However, the authors could perhaps mention some technology or method to help contextualize the type of solution that you have proposed in your particular experiment. And I am not sure if the term ‘practical project’ is appropriate in this context. Maybe you could evaluate the alternative of calling it a demo case, application case, case study...

1. Introduction

Lines 73-74: is not clear the meaning of sentence: “To neglect the obscure professional terms in the medical domain, we use the student profile scenario shown in Figure 1 to motivate our work”.

Lines 109-110: review if the expression “all walks of life” is appropriate.

Lines 119: I am wondering, is the question "what kind of knowledge graph can be reused?" the most interesting one in this context, or what kind of knowledge graph can I reuse to solve my need? and how can I know it. I guess all knowledge graph can be reused, but you have to find out which one best suits your needs.

Lines 125: “Figure 2 is an instance of schema extracted from Figure 1”: It does not appear to be an instance but rather a catch or excerpt. An instance would be course=2021, student=John, teacher=Joseph,… right?

Lines 126: If the authors are tackling the problem of structural differences, perhaps they should take a look at these articles: https://0-doi-org.brum.beds.ac.uk/10.1016/j.autcon.2020.103384 and http://ceur-ws.org/Vol-929/paper9.pdf

Lines 135-136: not sure about if ‘establishment’ is the correct word in this sentence.

Lines 142-142: “From the semantic definition point of view, these two standards have directions”: what do you mean with ‘directions’? Or at least, could you explain what they are?

I am going to skip the rest of the comments I had regarding the resto of the sections of the article to jump directly to the state-of-the-art section. I think there is a problem there that needs to be fixed first.

6. Estate of the art

The placement of this section at the end of the article is unusual and, in this case, it does not appear to be justified in any way. Therefore, the authors should move to the beginning after the introduction or as a subsection of it.

About the content, one of the key points of this article seems to be rooted in the question of how to query graphs schemas. However, this section begins by talking about data integration, as well as at the beginning of the second paragraph. The term 'data integration' only appears in this section and in the abstract, which suggests that the focus of the state of the art is misaligned with the experiment and, in general, with the approach followed in the rest of the article. This manifests an important problem in this article, which is how the authors position themselves and what is the starting point of what is examined in this article with respect to the state of the art to enable, this way, to determine the novelty of what is provided in this article.

7. Conclusions

It should be rewritten. This section is mainly focused on describing what has been presented in the article. It seems more like a summary than a conclusion. Readers would probably expect to read something about the results obtained and about future lines. This lack of expected result is also related to the state-of-the-art section: comments should be supported based on it.

Typographic errors:

Not detected. Just check for double spaces in the text.

References:

The number of references is enough.

Unless the journal has a section in the guide that enable it, the authors should add all the names of the authors in each of the references (with no ‘et al.’). And be sure to follow a journal-supported reference style with proper formatting.

Author Response

Dear reviewer,

Thank you for your comments. Your suggestions are very helpful to us. We have carefully revised the paper according to your comments, and the details are in the word file attached.

Best,

Authors

Author Response File: Author Response.docx

Reviewer 2 Report

I find the title attractive and the proposed work potentially interesting.

However, the paper is honestly confused and hard to understand (and much more to review/assess).

If knowledge graphs are underpinned by formal ontologies, ontology matching and the resulting semantic interoperability model should somehow support query on legacy datasets. Eventual quantifications/measures seem to depend on the effectiveness and accuracy of concept aligning firstly and on an effective management of the cross-knowledge defined by the integration process.

My feeling is that the work conducted has some potential but it is honestly not well communicated. I would suggest to completely re-write the paper accordingly. 

Author Response

Dear reviewer,

Thank you for your comments. Your suggestions are very helpful to us. We have carefully revised the paper according to your comments, and the details are in the word file attached.

Best,

Authors

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

The authors have made a remarkable effort to improve the quality of the manuscript. They have answered most of the comments indicated by the reviewer with success in most of cases. In particular, they have reorganized the sections of the article in a more logical order. However, there are issues that have not been adequately addressed. I am adding new comments and possible recommendations to improve the article in each of the sections.

As a general comment, edited parts seem like they should pass a revision in order to improve their quality.

Title

Now the title seems to be more appropriate according to the content of the article.

Abstract

The last sentence “The use of such criteria in the application case verifies the applicability” still seems a bit vague. For example, is the “use” or the “effectiveness” what has to be verified? Wouldn't it rather be something like “The effectiveness of the criteria obtained to evaluate the compatibility between Knowledge graphs and cross-domain queries has been demonstrated in a case study.”

1. Introduction

Figures 1 and 2 are in very low quality (in the original manuscript they were good).

2. Related work

Despite having reduced the data integration part, the term 'data integration' is too isolated with respect to the focus of this work and it would seem more appropriate if it were introduced in the previous section. The review of the related work should focus exclusively on analysing the state of the art of methods on semantic similarity criteria ¿right?

6. Experiment

Strange beginning of section. Hardware details usually appear towards the end.

7. Conclusions

Now the type of content in this section is much more focused on what a reader would expect to read in it. However, the level of quality of the wording and expressions used should be improved. For example, the expression in the last sentence is not usually appropriate for a scientific article. I suggest that readers seek the help of an experienced English writer if they cannot do it themselves.

Typographic errors:

Not detected.

References:

Although the authors have apparently applied a valid style format in the references, they should be reviewed as they contain errors. For example: “Seque Da , J.; Priyatna, F.; Villaz´On-Terrazas, B. Relational Database to RDF Mapping Patterns. In Proceedings of the International Conference on Ontology Patterns, 2013”, should be “Sequeda, J. F.; Priyatna, F.; Villazón-Terrazas, B. Relational Database to RDF Mapping Patterns. In Proceedings of the International Conference on Ontology Patterns, 2013”. There are more. I suggest that the authors take the references directly from indexed databases (Scopus, etc.) to avoid spelling problems. Also consider including the DOI in all those references that have it, since this facilitates access to the article by readers, and as indicated in the note provided on the left side of the first page of the document.

Author Response

Dear reviewer,

Thank you for your valuable suggestions. We have changed the manuscript according to your suggestions. The details are in the annex.

Bests,

Authors

Author Response File: Author Response.docx

Reviewer 2 Report

I appreciate the effort to improve the related work and the conclusions, as well as the answer to my comments by the authors.

Please note in the title that "criterion" refers to one, while "criteria" normally refer to more than one. Therefore, the "A" at the beginning of the title could be probably deleted.

Concerning the content of the paper, I strongly suggest to improve the presentation of the core part in the same way it has been done for related work and conclusions. I believe that is very important because I'm still unable to fully understand the contribution and I'm afraid many readers may experience similar issues.

Author Response

Dear reviewer,

Thank you for your valuable suggestions. We have changed the manuscript according to your suggestions. The details are in the annex.

Bests,

Authors

Author Response File: Author Response.docx

Round 3

Reviewer 1 Report

I think the authors have done their best to improve the article to try to bring it to as high a quality level as possible. They have answered most of the comments indicated by the reviewer with success in most of cases. However, there are still issues that have not been adequately addressed and there are parts of the text that in my opinion do not meet the level of quality required for this journal. Again, I would suggest that the authors ask for the help of a native / experienced English writer to help them improve the expressiveness of the writing.

References:

There are still errors. For example, “Horvat; Dunđer, M.; Lugović, I.; Sergej. Ontological heterogeneity as an obstacle for knowledge integration in the Semantic” in line 837. It should start as “Horvat, M.; Dunđer, I.; Lugović, S. …”. Or for example, the reference number 10, indicate the author's initials separated by periods: Vandenbussche, P.Y.; but in reference number 13, María, P.-V.; it goes with a hyphen. Please review it and put the correct format. I encourage authors to extract references from a common database (e.g., Scoups) using the same format, or to ensure that this is exactly the same format in all of them.

Reviewer 2 Report

The presentation of the paper has been significantly improved.

I'm therefore happy to endorse its publication.