Review Comment:
Since I did a first round of reviewing on this paper I stand by my original comments on Originality and Significance.
1) Originality - This paper aims to develop an ontology design pattern that defines exposure specifically in the context of environmental exposure, the health impacts on people, and the surrounding environmental influence of the exposure. Overall, this work is very interesting and much needed for several domains. The related section proves that there is a significant lack of reusable ontologies for exposure or any ontologies that use standardized definitions. The approach adopted by the authors to use terms/information extracted from scientific articles using NLP to drive the development of their ontology design pattern is great, although it might also be beneficial to see how standardized definitions for some of the concepts can be adopted.
2) Significance of the results - The development of such an ontology as this that represents the various aspects of exposure, including the agent or substance, the target, the environment of exposure, the duration and frequency of exposure, and the level or dose of exposure is significant in several contexts such as for health and safety, disaster management. This ontology can be useful to perform more sophisticated analyses and reasoning about the relationship between exposure and health outcomes. Furthermore, an ontology such as this can facilitate the development of computational tools and applications that rely on exposure data, such as exposure assessment models or decision support systems for risk management, ultimately contributing to a better understanding of the potential health impacts of environmental exposures.
3) Quality of writing – The authors have taken into consideration the feedback that was provided previously and addressed most issues. The readability and clarity of the paper are significantly better now. Having said that there are still some minor language issues and some modeling concerns that I have presented below:
General issues (formatting/language related):
1. Language inconsistencies: e.g., usage of both "modeled" and "modelled"
2. Throughout the paper the use of open quotation marks must be fixed (e.g., pg 2 line 28). If you are using latex editing software what you might want to use would be the backtick/left quote.
3. Be consistent with references. In some places, the authors’ names are used (e.g., pg 3 line 32, "... ontology by Zeshan and Mohamad [23]..."), whereas in others simply the reference number is used (e.g., pg 3 line 33, "... ontology by [24]..."). Adopting the latter format consistently might be better.
4. Pg 4 line 18: address --> addresses
5. Pg 4 line 43: sem-iautomated --> semi-automated
6. Pg 5 line 7: missing comma before "which"
7. Pg 8, line 32: including a reference for EPO:exposure class might be helpful
8. Pg 8, line 35: DCAT and PROV are introduced in the paper for the first time here, so mentioning their references would be good
9. Pg 11: Since Fig 2 seems to have a lot of white space around it, it might be helpful to draw out real-world examples (that are described in the text) against each exposure model
10. Pg 12 line 36: Should the concept "Environments" be "EnvironmentalFactor"? Both Fig. 3 and Fig. 4 only mention the latter as a class
11. Pg 13 line 20-21: missing parenthesis in the two axioms
12. Pg 14 line 35: what is the axiom 3 that is mentioned here? Perhaps numbering all the axioms might be helpful?
13. Pg 15 line 30: missing space after "These"
14. Pg 26-27 - It might be helpful to include references for the papers in these tables because its the reference numbers that are used in the Results section and not the paper names as mentioned in the tables here, so comparison is hard for the reader.
There are many more grammar/punctuation issues. Maybe using grammar software such as Grammarly (which is free) might be helpful in identifying and addressing these issues.
Modeling questions:
1. The notion of "temporal information" seems important for exposure, dose, and health risk. The authors talk about it quite a bit in the text on pg 11-13 (where they describe these concepts in detail). But the ontology does not include any temporal modeling, nor is it known how this information exists in any of the datasets (articles) used.
2. In page 3- the authors make the following claim "For every exposure there is a unique person who is exposed. Further the axioms in lines 20-21 state that an exposure is always caused by only one activity and the activity is always caused by only one person. This seems too restrictive IMO. Exposure can be due to a combination of several activities (e.g., dumping of toxic waste at a landfill site and a person's decision to live in proximity). In this example, there is no possible causal link between the dumping activity and a person's decision, and the exposure is rather a consequence of the two activities. Unless the implication is that for a given exposure, there is a primary, dominant, or most significant activity causing it. But is there even a possibility to determine this? Next, activities can be carried out by more than one person or even organizations (e.g., dumping of toxic waste at a landfill site by a waste-removal company). Likewise, an entire community (not simply an individual) can be exposed due to such an activity and not simply 1 person. This brings me to another question I have about Query 2 on Page 17 -- "Is the person who causes the activity always the one exposed". I cannot agree with this-- and there are so many real-world scenarios that may contradict this.
Overall I find this axiomatization too restrictive and maybe not a representation of many real-world scenarios.
3. I do not think the axiom for "Active" is doing what they say it's doing. It seems confusing.
4. Pg 19 line 4 - What do the authors mean by brute implementation? Manually or using a reasoner or inferencing within a graph database?
|