Knowledge Graphs and Explainable AI in Healthcare

Round 1

Reviewer 1 Report

Thank you for giving me the opportunity to review this manuscript. Authors did a good job in introducing and explaining the concept of knowledge graphs and existing work. Lately, I believe this topic is more relevant as many healthcare systems are contemplating to adapt to the usage of AI. However, inclusion of possible limitations of the XAI system in the manuscript would have been nicer!

I suggest the authors to carefully spell check the manuscript. For example: it is a bit confusing to the readers what the acronym XAI stands for when there are different abbreviations in Line 1 (EXplainable Artificial Intelligence) and Line 13(eXaplinable AI).

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Overall, this is a useful review, which may be further enhanced by considering the suggestions below: 

L10: "Intelligence" would be better.

L10/11: Citation needed.

L13: Should be: "eXplainable. 

L23/24: Citation needed. 

L59: Better as: "In a review paper [21], Zeng et al. summarized

commonly used...." 

L76: Should be "weight". 

L78: "Bayesian model" might benefit from a citation. 

L106: Should be: "multiple convolutional neural networks model (Multi-CNN)......."

L138: Should be: "Bayesian Personalized Ranking (BPR)....."

L147: Spell out OWL2 and give citation. 

L177: Might be better phrased for clarity. 

L236: Should be "...healthcare....."

L240: Should be: ".....information is transformed......."

L256: Awkward phrase: ".....system allows performing reasoning...."

L276: ".....noise......."

L277: Word choice? "......intruding...."

L295: Better as: "........created after......."

L313: "....noise......"

 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

 

This paper presents a state-of-the-art review of the role of knowledge graphs in explainable artificial intelligence models in healthcare. The topic is interesting and important considering knowledge graphs can provide human understandable explanations and valuable additional knowledge to XAI models. The authors categorized the knowledge graphs’ applications in the healthcare domain in pre-model, in-model and post-model XAI.

The suggestion is to add more tables/figures to better illustrate/summarize knowledge graph roles/applications in XAI in healthcare.

 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf