Reviewer has chosen to be Anonymous

Overall Impression: Good

Content:
Technical Quality of the paper: Good
Originality of the paper: Yes, but limited
Adequacy of the bibliography: Yes

Presentation:
Adequacy of the abstract: No
Introduction: background and motivation: Limited
Organization of the paper: Satisfactory
Level of English: Satisfactory
Overall presentation: Good

Detailed Comments:

On the Multiple roles of ontologies in Explainable AI.
Roberto Confalonieri, Giancarlo Guizzardi

This paper gives an overview of the different roles of ontologies in Explainable AI.
They consider three roles for ontologies: (i) reference modelling, (ii) common-sense reasoning, (iii) knowledge refinement and complexity management. Along those roles, they discuss existing works. A very valuable overview of XAI methods that exploit ontologies.

The paper gives indeed a nice overview. It would be nice to make the different categories more fine grained such methods can be more easily compared. However I do understand that it is also difficult.

Main points:
(1) The most important suggestion is to include a paragraph in the introduction that explains why this explainability paper fits well in the new journal of Neurosymbolic. Include also adding a sentence in the abstract. In my opinion the paper fits the journal, because explainability methods are discussed that explain a black box (ML) by exploiting knowledge (reasoning).
Include a more prominent link to neurosymbolic in your "summary" section as well.

(2) The authors summarise a number of papers. Would it be possible to add motivation to the selection of papers?

(3) it would be nice to spend a paragraph on each challenge, explain and discuss the three challenges in more detail.
include a more prominent link to neurosymbolic in your "summary" section.

Smaller remarks:
page 2, row 48: “AI systems” do you mean here ML systems?
page 8, row 12: “depends on the specific requirements of each application”. A concrete example might be helpful.

Comments on the scores:
Abstract: No —> I would like to see the neurosymbolic link
Introduction: Limited —> I would like to see the neurosymbolic link

Reviewer has chosen not to be Anonymous

Overall Impression: Excellent

Content:
Technical Quality of the paper: Excellent
Originality of the paper: Yes
Adequacy of the bibliography: Yes

Presentation:
Adequacy of the abstract: Yes
Introduction: background and motivation: Good
Organization of the paper: Satisfactory
Level of English: Satisfactory
Overall presentation: Excellent

Detailed Comments:

This paper proposes an interesting overarching classification of approaches that use symbolic knowledge (in the form of ontologies) for ensuring explianability of AI systems, thus discussing an interesting and timely topic in the area of neurosymbolic systems.

The material is thoroughly researched and points to a wide range of papers from various domains. The summaries of various representative works are easy to understand and supported by illustrative examples.

The paper concludes with a number of challenges in using ontologies for explainable AI.

The paper is suitable for publication as is. Possible, but not mandatory improvements could be:

(1) better highlight the relation and relevance to the neurosymbolic field in the abstract/introduction. While this relevance is clear once the paper is read, making it more prominent at the beginning of the paper would be important especially given that this paper will appear in an inaugural issue of the NAI journal.

(2) section 6 which discusses challenges could be more structured (e.g., include headings/bulleted list for each challenge) and potentially extended with additional challenges - in fact, we can assume that such understanding of challenges in an emerging field is of high interest to readers of the planned inaugural issue. Consider also renaming it to "Open Challenges". Also, the summary of the challenges at the end of Section 7 does not fully overlap with what is discussed in Section 6, so a tighter alignment in content should be attempted.