Development of an Indoor Space Semantic Model and Its Implementation as an IndoorGML Extension†
Round 1
Reviewer 1 Report
In this publication was defined the basic objects necessary in the semantic indoor model. The model includes one floor. In this model lacks stairs, elevators. These objects would allow connecting two floors. In the model, other usability objects should be added, eg a balcony, terrace, porch, vestibule. The next model version should be connected with the gas, water, energy and other facilities. The semantic description of objects should be extended with new attributes and dictionaries.
Author Response
Response to Reviewer 1 comments
Point 1 : In this publication was defined the basic objects necessary in the semantic indoor model. The model includes one floor. In this model lacks stairs, elevators. These objects would allow connecting two floors.
Response 1 : Initially the model was meant for describing spaces in a single floor though the model was conceptualised for multi-level buildings. As suggested by the reviewer, we have now included the vertical spaces/connectors to make the paper more comprehensive. (relevant Sections are 3.2.4 and 3.2.5).
Point 2 : In the model, other usability objects should be added, eg a balcony, terrace, porch, vestibule.
Response 2 : Owing to the heterogeneity in the nomenclature used by different architects, a macro level classification model is proposed in the paper, based on enclosure or containment. The multiple spaces and their mapping to generic space types, based on the classification model, has been provided in the Table 1 and described under the different space classes. Hope this clarifies how the various spaces are handled in the paper. Also refer to Section 2.1 for a full classification of indoor space environment.
Point 3 : The next model version should be connected with the gas, water, energy and other facilities. The semantic description of objects should be extended with new attributes and dictionaries.
Response 3 : As suggested by the reviewer we are also planning to do this in the future as the next version of the work. But at this stage this is not part of the current scope of the paper. Having said that, the current model, a generic one, is an extendible model which can include objects depending upon the application where the model needs to be utilized.
Reviewer 2 Report
The article presents many correct assumptions and views of the authors. The article touches on important issues and is interesting from this point of view. The introduction to the article is well-written and the reason for the research is well-motivated.
However, the described research does not solve the problems discussed deep enough.
Proposed solutions is not too innovative. This issue was discussed in the literature and more advanced solutions have already been proposed.
The proposed models are too simplistic and do not take into account the full specifics of space inside buildings.
The authors do not state, why they do not distinguish e.g. such subclasses or classes as Elevators and Stairs, and use only one of the Corridor class. It seems that the authors omitted the issue of presenting the door connecting the corridor with the exterior of the building. They define that "A the door provides a passage and connects two different spaces to each other".
Lack of clear justifications for the proposed models.
There is no reference to the literature items that discuss these problems much deeper.
This proposal of class diagram (fig. 7) requires a longer discussion. For example, it is not clear why the Hall class was not associated with other classes (eg aggregation association), but the SemanticSpaceType data type was used. Why the Hall class is on the same level in the model as the Corridor, Opening etc.? Perhaps this is not a good solution.
Appendix, it not seems to be needed in so wide extent
In conclusion, in my opinion, the proposed solution is therefore not comprehensive, is too superficial and the proposals presented are not fully explained. There is no clearly described research methodology. There is no verification of the proposed models.
Author Response
However, the described research does not solve the problems discussed deep enough.Proposed solutions is not too innovative. This issue was discussed in the literature and more advanced solutions have already been proposed.
The proposed models are too simplistic and do not take into account the full specifics of space inside buildings.
Point: The
authors do not state, why they do not distinguish e.g. such subclasses
or classes as Elevators and Stairs, and use only one of the Corridor
class.
Response: The classes 'Stairwell and Lift well' have been added to model the respective spaces.
Point: It seems that the authors omitted the issue of presenting the door connecting the corridor with the exterior of the building. They define that "A the door provides a passage and connects two different spaces to each other".
Lack of clear justifications for the proposed models.
There is no reference to the literature items that discuss these problems much deeper.
Point: This proposal of class diagram (fig. 7) requires a longer discussion. For example, it is not clear why the Hall class was not associated with other classes (eg aggregation association), but the SemanticSpaceType data type was used. Why the Hall class is on the same level in the model as the Corridor, Opening etc.? Perhaps this is not a good solution.
Point: Appendix, it not seems to be needed in so wide extent
In conclusion, in my opinion, the proposed solution is therefore not comprehensive, is too superficial and the proposals presented are not fully explained. There is no clearly described research methodology. There is no verification of the proposed models.
Reviewer 3 Report
The scales of diagrams on figures 7 and 8 are different. The diagram in figure 7 is much smaller and less readable than diagram 8. I suggest redesigning the UML diagram in figure 7 to make it fit the paper width with the same scale as figure 8. There is also no reference to green and yellow colors of features on each diagram. The description that is available in lines 292-302 should mention it.
Chapter 3 is very similar to chapter 3 of the "A SEMANTIC MODEL TO DEFINE INDOOR SPACE IN CONTEXT OF EMERGENCY EVACUATION" paper (by Nishith Maheshwari, and K. S. Rajan). The work is referenced in this publication, but I feel like there is no clear reference to this work in text inside chapter 3.
The work is interesting and clearly prepared. Some minor spelling errors can be found. For example: line 28: it is line 38: "have" instead of "has", "their" instead of "its". Or "The format" instead od "The formats". line 47: using domain-specific data line 57: make line 60: well-defined etc.
Author Response
Response to Reviewer 3 comments
Point 1 : The scales of diagrams on figures 7 and 8 are different. The diagram in figure 7 is much smaller and less readable than diagram 8. I suggest redesigning the UML diagram in figure 7 to make it fit the paper width with the same scale as figure 8. There is also no reference to green and yellow colors of features on each diagram. The description that is available in lines 292-302 should mention it.
Response 1 : As suggested by the reviewer, the figures have been modified to match the same scale. The features in green color belong to the IndoorGML core module and the features in yellow are the ones defined in our model. The text has been modified to mention the same (Line number 434).
Point 2 : Chapter 3 is very similar to chapter 3 of the "A SEMANTIC MODEL TO DEFINE INDOOR SPACE IN CONTEXT OF EMERGENCY EVACUATION" paper (by Nishith Maheshwari, and K. S. Rajan). The work is referenced in this publication, but I feel like there is no clear reference to this work in text inside chapter 3.
Response 2: While the space model proposed in the previous paper referred to a very basic framework, this paper substantially builds on that and proposes a more exhaustive approach in trying to define/classify the indoor spaces, its ontological framework and further a data model in which it can be implemented.
As suggested by the reviewer, a clear reference to the previous work with the commonality and the differences have been given in the beginning of Section 3 (Line numbers 261-264).
Point 3 : The work is interesting and clearly prepared. Some minor spelling errors can be found.
Response 3: The mentioned grammatical and spelling errors by the reviewer have been rectified.
line 28: it is
Line 29
line 38: "have" instead of "has", "their" instead of "its". Or "The format" instead od "The formats".
Line 39
line 47: using domain-specific data
Line 49
line 57: make
Line 58
line 60: well-defined
Line 61
Round 2
Reviewer 2 Report
The authors took into account some of the reviewer's comments and significantly expanded the content of the article.
In the new parts of the article, authors should check the language verification of, for example, commas.
In rows 314, 330, 363, 373, 388, the attribute "Floor" was added.
This is ok, but it must be noted that there are some modern buildings where the number of the floor can not be clearly defined.
Although not everything has been corrected, I think that the article can be released for publication
