Does Time Smoothen Space? Implications for Space-Time Representation
Round 1
Reviewer 1 Report
This paper has studied the issue of space, time and continuity through granularity, completeness, uncertainty, scale, precision, ontology and topology in spatial and spatio-temporal data and analysis to investigate the space and time interaction in GIS. A number of GIS analyses including intersect and overlay, visibility analysis, graph partition and granulation, path analysis, …..have been studied to support the idea. The paper consists of two parts. Part 1 has discussed reasoning why sphere packing limits suggest that a continuum cannot be completely addressed. Part 2 has explained that paradox by creating a smooth space from discrete topological elements to examine the relationship between space and time for its representation. Face Centered Cubic Packing is then proposed as a practical approach to model, the issue of incompleteness in spatio-temporal representation. It is shown that the emergent dimension of “Lattice Communication Time” (LCT) is continuous. Each path between two nodes on the graph can thus be mapped as a vector between points in this continuous, as a hyperspace. Although no exact response to the questions raised on the relationships between space and time dimension decencies and correlation. The paper has tried to provide some paths to verify the conceptual views on the subject. The reasoning seems justifiable and the paper is publishable provided the following points can be considered in the revised version:
1- How the idea of this paper matches with modifiable areal unit problem (MAUP) and modifiable temporal unit problem (MTUP)?
2- What type of graph is considered in this paper among directed graph, convex graph, concave graph, 3D graph, ….?
3- “Discreet” should be replaced with “Discrete”.
4- There are a number of numbering inconsistencies for sections and sub sections in the paper. For example, 1.1.1.1 has been used for Intersect and Overlay, Visibility Analysis, Graph Partition and Granulation, Path Analysis, Measurement Precision and Topological Vagueness, Building dimensions, Dimensions, and Precision in projected space.
5- Ohori et al. [81] is not the recent study to argue that ‘true’ 4D requires time be topologically integrated with space. Some previous studies such as Neysani Samany et al. (2019), Developing FIA5 to FSTPR25 for modeling spatio‑temporal relevancy in context‑aware wayfnding systems, Journal of Ambient Intelligence and Humanized Computing, https://doi.org/10.1007/s12652-019-01287-1, have already pointed out the necessity for a harmonized space-time topological presentation and analyses.
6- What do 3A, 3B and 3C mean in the caption of Figure 6?
7- Introduction lacks research background, necessity, scenario, questions, hypotheses, assumptions, and major intended innovations and contributions.
8- The paper has not provided any quality control and quality assessment for the proposed claims.
9- Why Euclidian space is considered and how about other spatio-temporal spaces?
10- The paper lacks a clear innovations and contribution, however, presenting the existing knowledge on space and time continuity, uncertainty and granularity and providing some GIS examples on the need to care about granularity in space and time could be an interesting topic to be published.
Author Response
General responses to points from several reviewers:
My thanks to the reviewers for the supportive and detailed comments. In response the article has been streamlined by moving from the two part structure to sequential sections and cutting the more tangential material. Overall the article is over 1200 words and 44 references shorter (two references are added at reviewers suggestion) and several of the footnotes have also been cut. The structure of the paper has been changed as follows:
A new introduction has been written, the first sentence of which states the initial hypothesis on addressing space. The second paragraph explains how that lead to the title question and the closing paragraph clearly states the two key messages of the paper.
Although the journal template does not fit a theoretical paper comfortably, the new draft does attempt to follow it more closely:
Section 1: Research Background. The theoretical introduction has been cut substantially, allowing the reader to reach the applied examples of granularity sooner. The range of examples has not been reduced as each relates to a different aspect of granularity, but it leads step by step toward lattices and so directly to the next section.
Section 2 : I have re-written the text to be more methodical in style and have shortened the interim conclusions. If the editor would prefer to see the term method in the section titles for sections 2 and 3 this would also be acceptable.
Section 3: The review of Space-time representation as been condensed. Formats where it was part of the review in section 1 were attempted but interfered with the flow from section 1 to 2.
3.2 Some elements have been cut to improve readability, including a figure and some quotations but the need to be complete, precise and concise does limit options. I am not sure if the supplemental material was supplied to reviewers with the paper, but it does provide a plain language (if less precise) explanation of the section.
Section 4: Discussion. The general comparison between LCT and Peuquet's space time characteristics has been cut to half a page and the two broad questions removed in order to expedite the discussion straight to the section on FCCP.
Section 5: Conclusion. Reduced in length by removing some lesser points. MAUP and MUTP are now connected to Bitter and Smiths object discontinuity and Wolfs 'possible parts' as the penultimate point in the paper. The final point is that FCCP addresses that issue. The theoretical and practical contributions noted in the new introduction are thus now the concluding two points of the paper.
Reference numbers are now in order of first use and section numbers are sequential (NB section numbers were sequential in the original document, if they are again all numbered 1.111 etc. there must be some technical issue post submission).
Specific responses to Reviewer 1
>> How the idea of this paper matches with modifiable areal unit problem (MAUP) and modifiable temporal unit problem (MTUP)?
MAUP is referred on p.2 as an implication of Wolf's thoughts. MTUP is now referenced there also as well as in the conclusion where the same is discussed in relation to time. I would like to talk about this aspect more but given the general consensus of reviewers to streamline the paper and its references I cannot justify this.
>>What type of graph is considered in this paper among directed graph, convex graph, concave graph, 3D graph, ….?
The following has been added on p.8 (269) "un-directed and without pre-assigned global characteristics (such as dimensions or convexity) ensures no implicit background. "
>> “Discreet” should be replaced with “Discrete”.
Done
>>There are a number of numbering inconsistencies for sections and sub sections in the paper. For example, 1.1.1.1 has been used for Intersect and Overlay, Visibility Analysis, Graph Partition and Granulation, Path Analysis, Measurement Precision and Topological Vagueness, Building dimensions, Dimensions, and Precision in projected space.
This error was introduced after submission. I have corrected it back again. The sections are also substantially simplified.
>>Ohori et al. [81] is not the recent study to argue that ‘true’ 4D requires time be topologically integrated with space. Some previous studies such as Neysani Samany et al. (2019), Developing FIA5 to FSTPR25 for modeling spatio‑temporal relevancy in context‑aware wayfnding systems, Journal of Ambient Intelligence and Humanized Computing, https://doi.org/10.1007/s12652-019-01287-1, have already pointed out the necessity for a harmonized space-time topological presentation and analyses.
Thankyou for highlighting this interesting reference, fuzzy spatio-temporal prisms are now cited in the review section.
>> What do 3A, 3B and 3C mean in the caption of Figure 6? NB NOW FIG 5
It now reads A) B) C) referring to the three parts of the figure respectively.
>> Introduction lacks research background, necessity, scenario, questions, hypotheses, assumptions, and major intended innovations and contributions.
The first two paragraphs now address these directly.
>>The paper has not provided any quality control and quality assessment for the proposed claims.
Thanks for highlighting this. I suspect there many ways to compare which will be contingent on application but I have now suggested a general test on line 569.
>>Why Euclidian space is considered and how about other spatio-temporal spaces?
Added to section 2 ". Since the conjecture and Hales Proof of it are confined to Euclidean space, any implications must be restricted to it. But within that common analytical arena...".
>>The paper lacks a clear innovations and contribution, however, presenting the existing knowledge on space and time continuity, uncertainty and granularity and providing some GIS examples on the need to care about granularity in space and time could be an interesting topic to be published.
The contributions are now specified in the first two paragraphs of the introduction. Beyond these I believe the discussion takes a novel angle on the issues which is itself of value.
Reviewer 2 Report
I found this a very difficult paper to review: it is bursting with ideas and has over 130 references, but I struggling to find the thread running through the paper. Perhaps it is the rather wordy style, or the very theoretical subject matter but it was, for me, quite a difficult paper to follow.
I would urge the author to think about the key messages they wish the reader to take away from the paper and to bring these front and centre. The structure of the paper, with two semi-connected parts doesn't help in this regard. I don't see the utility of splitting the paper in two like this?
The introduction gives a good background the problem and sets out the concept of granularity nicely, but the actual applied example was, for me, hard to follow. The author should bring the applied example forward in the paper to explain the purpose of the rather long and wide-ranging review of the history of spatial representation and granularity.
I think there is the nugget of some interesting theoretical work in this paper which, with some judicious editing and re-ordering, could make a good contribution to the literature.
In the version I reviewed, there is something wrong with the formatting, with all sections labelled as 1.1.1.1.
Some minor typos are present throughout, so it could do with a proofread. E.g:
Wolf's
Line of site
Discreet/discrete
Cannot be validated in *the* field
Author Response
General responses to points from several reviewers:
My thanks to the reviewers for the supportive and detailed comments. In response the article has been streamlined by moving from the two part structure to sequential sections and cutting the more tangential material. Overall the article is over 1200 words and 44 references shorter (two references are added at reviewers suggestion) and several of the footnotes have also been cut. The structure of the paper has been changed as follows:
A new introduction has been written, the first sentence of which states the initial hypothesis on addressing space. The second paragraph explains how that lead to the title question and the closing paragraph clearly states the two key messages of the paper.
Although the journal template does not fit a theoretical paper comfortably, the new draft does attempt to follow it more closely:
Section 1: Research Background. The theoretical introduction has been cut substantially, allowing the reader to reach the applied examples of granularity sooner. The range of examples has not been reduced as each relates to a different aspect of granularity, but it leads step by step toward lattices and so directly to the next section.
Section 2 : I have re-written the text to be more methodical in style and have shortened the interim conclusions. If the editor would prefer to see the term method in the section titles for sections 2 and 3 this would also be acceptable.
Section 3: The review of Space-time representation as been condensed. Formats where it was part of the review in section 1 were attempted but interfered with the flow from section 1 to 2.
3.2 Some elements have been cut including a figure and some quotations but the need to be complete, precise and concise does limit options. I am not sure if the supplemental material was supplied to reviewers with the paper, but it does provide a plain language (if less precise) explanation of the section.
Section 4: Discussion. The general comparison between LCT and Peuquet's space time characteristics has been cut to half a page and the two broad questions removed in order to expedite the discussion straight to the section on FCCP.
Section 5: Conclusion. Reduced in length by removing some lesser points. MAUP and MUTP are now connected to Bitter and Smiths object discontinuity and Wolfs 'possible parts' as the penultimate point in the paper. The final point is that FCCP addresses that issue. The theoretical and practical contributions noted in the new introduction are thus now the concluding two points of the paper.
Reference numbers are now in order of first use and section numbers are sequential (NB section numbers were sequential in the original document, if they are again all numbered 1.111 etc. there must be some technical issue post submission).
Specific responses to Reviewer 2
>>think about the key messages they wish the reader to take away from the >>paper and to bring these front and centre.
This is now speficied in the opening paragraphs and returned to in the conclusion.
>> structure of the paper, with two semi-connected parts doesn't help in this >>regard. I don't see the utility of splitting the paper in two like this?
The parts have been removed. It is now a single paper of four sections and a conclusion.
>>long and wide-ranging review of the history of spatial representation and granularity.
The theory section has been cut to a single page. One paragraph may still be argued to fall more into history of philosophy, but its relevant and unlikely to appear in typical GIS literature lists.
>> Bring the applied examples earlier on.
The introductory theory has been cut to a single page, which bring the examples near to the beginning without preceeding general theory.
>> "Minor edits: Wolf's
Line of site
Discreet/discrete
Cannot be validated in *the* field"
These have been changed and the rest of the paper checked manually and via MS Word for futher such issues. I am happy to pay for a professional language check to catch such errors in the accepted version if any remain.
Reviewer 3 Report
The format of the paper has a big problem, it is suggested that the author carefully compare the template to modify.
The sequence of references is very confusing.
The formula in the paper and the explanation of the formula are not standard. It looks like the formula is just a pasted picture.
The sequence number and structure of paragraphs are very confusing.
Author Response
General responses to points from several reviewers:
My thanks to the reviewers for the supportive and detailed comments. In response the article has been streamlined by moving from the two part structure to sequential sections and cutting the more tangential material. Overall the article is over 1200 words and 44 references shorter (two references are added at reviewers suggestion) and several of the footnotes have also been cut. The structure of the paper has been changed as follows:
A new introduction has been written, the first sentence of which states the initial hypothesis on addressing space. The second paragraph explains how that lead to the title question and the closing paragraph clearly states the two key messages of the paper.
Although the journal template does not fit a theoretical paper comfortably, the new draft does attempt to follow it more closely:
Section 1: Research Background. The theoretical introduction has been cut substantially, allowing the reader to reach the applied examples of granularity sooner. The range of examples has not been reduced as each relates to a different aspect of granularity, but it leads step by step toward lattices and so directly to the next section.
Section 2 : I have re-written the text to be more methodical in style and have shortened the interim conclusions. If the editor would prefer to see the term method in the section titles for sections 2 and 3 this would also be acceptable.
Section 3: The review of Space-time representation as been condensed. Formats where it was part of the review in section 1 were attempted but interfered with the flow from section 1 to 2.
3.2 Some elements have been cut including a figure and some quotations but the need to be complete, precise and concise does limit options. I am not sure if the supplemental material was supplied to reviewers with the paper, but it does provide a plain language (if less precise) explanation of the section.
Section 4: Discussion. The general comparison between LCT and Peuquet's space time characteristics has been cut to half a page and the two broad questions removed in order to expedite the discussion straight to the section on FCCP.
Section 5: Conclusion. Reduced in length by removing some lesser points. MAUP and MUTP are now connected to Bitter and Smiths object discontinuity and Wolfs 'possible parts' as the penultimate point in the paper. The final point is that FCCP addresses that issue. The theoretical and practical contributions noted in the new introduction are thus now the concluding two points of the paper.
Reference numbers are now in order of first use and section numbers are sequential (NB section numbers were sequential in the original document, if they are again all numbered 1.111 etc. there must be some technical issue post submission).
Specific responses to Reviewer 3
>>The format of the paper has a big problem, it is suggested that the author >>carefully compare the template to modify.
The format has been simplified substantially and does now follow the IJGI template.
>>The sequence of references is very confusing.
References are now numbered in order of first use.
>>The formula in the paper and the explanation of the formula are not >>standard.
The correctness of the formulas and their explanations has been checked with a consultant who holds a PhD in astrophysics. If there is something specific which is incorrect or could be better phrased please suggest it.
>>It looks like the formula is just a pasted picture.
The journal has been supplied with the original Latex version. Embedding this in Word seems to be not straight forward.
>>The sequence number and structure of paragraphs are very confusing.
The numbering issue was due to a technical error, post submission. The overall structure is also now substantially different.
Reviewer 4 Report
It is after a long time that I have read such a theoretical paper about Space-Time GIScience (the extensive references to Pequet, Langran, Egenhofer, Shortridge et al. is a throwback to 20 years ago, when such topics were the forefront of the field). It is interesting to see the author present and defend a conjecture (by citing Hales proof of Keppler's conjecture on the hard packing of spheres suggest it may not be able to address spheres completely), alongwith the use of Face Centered Cubing packing as a means for creating space time topology.
However, the innovativeness of the idea is lost in the jargon-filled writing that seems to have been haphazardly put together. Overall, the paper can be significantly simplified for the average reader. For someone like myself, who has worked on spatio-temporal granularity (and yes, of the event-based kind that always depended on a jarring breakup of time into fragments), this paper was a difficult read at best. It brings in too much theory with not much clarity. For example, why is it important to talk about Egenhofer's 9-intersection model when the model proposed isn't really about the relationship between 2 objects in space?
Moreover, it seems like the paper has been cobbled together from several different ideas and material written over different periods of time, so it does not form a cohesive whole. Note that even the citations are not in the correct order (the first citation is #82). In its current form, it is terribly unintelligible to the average reader.
I would therefore recommend a re-write keeping the flow of the material presented and clarity required for the reader in mind. Also, the number of reference can be cut down to keep only the relevant ones. Basically, this paper needs a focussed explaination of the new proposed FCC structure for space-time topology. Moreover, some post-hoc comparisons to existing models (e.g., some of the event-based spatio-temporal or composite models) would be helpful.
Overall, however, I would look forward to seeing some of the ideas presented, if laid out in a more readable format, in print.
Author Response
General responses to points from several reviewers:
My thanks to the reviewers for the supportive and detailed comments. In response the article has been streamlined by moving from the two part structure to sequential sections and cutting the more tangential material. Overall the article is over 1200 words and 44 references shorter (two references are added at reviewers suggestion) and several of the footnotes have also been cut. The structure of the paper has been changed as follows:
A new introduction has been written, the first sentence of which states the initial hypothesis on addressing space. The second paragraph explains how that lead to the title question and the closing paragraph clearly states the two key messages of the paper.
Although the journal template does not fit a theoretical paper comfortably, the new draft does attempt to follow it more closely:
Section 1: Research Background. The theoretical introduction has been cut substantially, allowing the reader to reach the applied examples of granularity sooner. The range of examples has not been reduced as each relates to a different aspect of granularity, but it leads step by step toward lattices and so directly to the next section.
Section 2 : I have re-written the text to be more methodical in style and have shortened the interim conclusions. If the editor would prefer to see the term method in the section titles for sections 2 and 3 this would also be acceptable.
Section 3: The review of Space-time representation as been condensed. Formats where it was part of the review in section 1 were attempted but interfered with the flow from section 1 to 2.
3.2 Some elements have been cut including a figure and some quotations but the need to be complete, precise and concise does limit options. I am not sure if the supplemental material was supplied to reviewers with the paper, but it does provide a plain language (if less precise) explanation of the section.
Section 4: Discussion. The general comparison between LCT and Peuquet's space time characteristics has been cut to half a page and the two broad questions removed in order to expedite the discussion straight to the section on FCCP.
Section 5: Conclusion. Reduced in length by removing some lesser points. MAUP and MUTP are now connected to Bitter and Smiths object discontinuity and Wolfs 'possible parts' as the penultimate point in the paper. The final point is that FCCP addresses that issue. The theoretical and practical contributions noted in the new introduction are thus now the concluding two points of the paper.
Reference numbers are now in order of first use and section numbers are sequential (NB section numbers were sequential in the original document, if they are again all numbered 1.111 etc. there must be some technical issue post submission).
Specific responses to Reviewer 4
>>Why is it important to talk about Egenhofer's 9-intersection model when the model proposed isn't really about the relationship between 2 objects in space
The relevance was that it considered the empty set, which is essentially the interstitial space this paper concerns itself with (that which is not addressed). However, as I have cut down the theory part substantially the reference no longer appears.
>>recommend a re-write keeping the flow of the material presented and clarity required for the reader in mind.
The paper as been re-cut into four sequential sections bookended by clearly liniked introduction and conclusion. Much contextual, but arguably tangential, material has been removed.
>>the number of reference can be cut down to keep only the relevant ones
The total number of references has been cut by 44.
>> Re Write to focus on FCCP
The Discussion section has been re-written to move more directly from the LCT to FFCP. The conclusions have also been streamlined. Over all the paper has been shortened considerably.
>>some post-hoc comparisons to existing models (e.g., some of the event-based spatio-temporal or composite models) would be helpful
It will be important to implement FCCP and test it against other models, how to do so will be application dependent and is thus future work. However a general metric has been suggested on line 569.
>>Overall, however, I would look forward to seeing some of the ideas >presented, if laid out in a more readable format, in print.
Thankyou.
Round 2
Reviewer 3 Report
i have no other suggestion
