An Intelligent Visualisation Tool to Analyse the Sustainability of Road Transportation

Round 1

Reviewer 1 Report

• The authors need to show labels for the x-, y- and z- axis in Figure 2 – 6, and clearly indicate what they are. 
• Page 14, Line 76-79, “… [13] determine that emissions in light vehicles… This variable is given special relevance in this study.” 1) throughout the manuscript, I couldn’t see how this “relevance” was addressed. 2) Vehicles used in freight transport are usually heavy trucks, so how is “light vehicles” given special relevance to the study? 
• As mentioned in Section 2.2, Line 247-260, the main different of proposed method and the existing HUEP is the use of KPAC or SM instead of PAC/MLHL/CMLHL, thus, I believe one of the central questions is how the use of KPAC and SM in HUEP is compared with the existing HUEP. Unfraternally, this is not addressed anywhere in the manuscript. 
• Section 2.2, Line 234-237, “…, 3 EPP methods (mainly based on ANN) were put forward, commonly known as PCA, MLHL and CMLHL.” The authors need to know that ANN is different from the PCA, MLHL and CMLHL in many aspects.
• The authors should add a methodology section to show the details of the configuration of the experiments so that the results could be reproduced, rather than only giving a very brief list (Page 14, Line 238 – 246).  
• Section 2.2, Line 245 – 246, “added sustainability data”; Section 3.1, “additional information on Fleet Age…”. Was the “Fleet Age” considered as “sustainability data”? 
• Page 13, Line 279, “Since it is not possible to include all the results obtained…” I understand there may be page limitation, however, the full set of the experimental results should be attached, at least, in appendices, in order for readers to evaluate the validity of the methods proposed in the manuscript. 
• Section 3.1, Figure 2, “After having selected KPCA as the EPP model that offers the best projections for the data analysed,…” If the KPCA offers the best projections, what is the point for proposing the use of SM in Section 2.2?  
• Datasets are explained in Section 2.1, however, the authors didn’t explain which attributes were used for clustering, in all the figures (Figure 2 – 6).
• It is very hard to interpret the Figures. For example, the yellow dashed line in Figure 3, how one could suggest that it is useful when determining the phases of progression of the age series (I don’t think the position of the projection on each dimension suggests anything in the figure)?
• I believe that pollution reduction is another central topic of this work (as listed in the keywords), but I don’t see any analysis or study regarding pollution reduction throughout the manuscript.
• It is hard for me to see how the conclusions reflect the study and analysis conducted in the main body of the manuscript. 

Author Response

Please see the attached file with the responses.

Author Response File: Author Response.pdf

Reviewer 2 Report

The major concern with this article is that the novelty is unclear.

 

The study uses HUEPs, but does not clearly specify what was improved compared to the [29] “Á. Herrero, A. Jiménez, and S. Bayraktar, “Hybrid Unsupervised Exploratory Plots: A Case Study of Analysing Foreign Direct Investment,” Complexity, vol. 2019, p. 6271017, 2019”

 

Apart from this, the work does not clearly state what is the value-add of the presented charts, what are the new significant insights discovered with the applied visualization techniques.

 

In addition, the conclusion “…provides better results…” (section 4) is not grounded on state-of-the-art metrics and seems to be highly subjective.

 

Related work regarding the visualization techniques is almost absent.

Author Response

Please see the attached file with the responses.

Author Response File: Author Response.pdf

Reviewer 3 Report

Interesting paper. In the reviewed paper, in order to analyze the sustainability of this activity, the Authors proposed the application of novel Artificial Intelligence techniques (more specifically, Machine Learning). In this research, the use of Hybrid Unsupervised Exploratory Plots is broadened with new Exploratory Projection Pursuit techniques. These, together with clustering techniques, form an intelligent visualization tool that allows knowledge to be obtained from a previously unknown dataset. The proposal was tested by the Authors with a large dataset from the official survey for road transport in Spain, which was conducted over a period of 7 years. The results obtained are interesting and provide encouraging evidence for the use of this tool as a means of intelligent analysis on the subject of developments in the sustainability of road transportation. In my opinion,  please take into account the following remarks:

• on the figure "Figure 1. Transport fleet age data series, divided into quartiles and related to the phases determined for the Great Depression" the names and units of axis "x" and axis "y" should be added. This remark has dedicated all figures with axis "x" and axis "y" in the paper text,
• the name of subsection 2.2. is "HUEPs".All acronyms given in the paper text the first time should be first explained. So, in this case, it is recommended to give the full name of subsection 2.2. i.e. "Hybrid Unsupervised Exploratory Plots",
• the Authors proposed the use of the intelligent visualization tool to analyze the sustainable development of road transport and in the article a significant part of the content is devoted to the description and characteristics of this tool. That is good, but unfortunately, the road transport part is insufficiently described in the paper and needs to be expanded. The data used in the paper text to the analysis relating to road transport of goods, so the literature review in this area should be expanded, indicating, in general, the issues of international freight transport, or cargo securing techniques, or the issues of transporting oversized cargo, e.g.
  "International road cargo transport in Poland and other EU countries" doi 10.20858 / SJSUTST.2021.111.8; "Essential techniques for fastening loads in road transport", doi 10.20858 / sjsutst.2021.110.8;
  The Transport of Oversized Cargoes from the Perspective of Sustainable Transport Infrastructure in Cities ", doi.org/10.3390/su13105524; "Oversize cargo transport in road transport - Problems and issues" doi 10.20858 / sjsutst.2020.108.12.
  One short paragraph in the Introduction and previous work section will be enough,
• lines from 240 to 246: in order to describe each original x vector (from the input space), the Authors used the numbers 1,2,3,4. It is recommended to replace these numbers with bullet items as is required by the Sustainability journal paper template,
• in some places paper text is not formatted according to the Sustainability journal requirements. It should be improved,
• the meaning of the letter "Q" used in Table 2 is not explained. It should be improved,
• the meaning of the yellow line on figure 3 should be explained. The meaning of this yellow line is explained in the paper text, but it also should be explained in a legend on the figure,
• there is a lack of discussion section in which the Authors present a comprehensive discussion about the obtained results. It should be added.

Author Response

Please see the attached file with the responses.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

In this revised manuscript, the authors have addressed most of the issues from the previous review. However, there are still some central questions that need to be further clarified. 

 

1.  It is common that the performance of a selected analysis algorithm is dependent on the data to be analysed. I believe the authors agree with this, Page 8, Line 319 – 320, “After having selected KPCA as the EPP model that offers the best projections for THE data analysed,”. If the authors used a different dataset from the one that was used to evaluate the original HUEP, then the contribution of the authors would be limited to this very particular combination of the dataset and KPCA. If this is the case, the authors need to clearly state the limitation of the work.  
2. In the Results section, it is hard to see what we can conclude and what we can discover from these plots? For example, in Figure 3, as the authors state that “The yellow line is associated to the temporal progression of data, From a practical point of view, the graph shows that the visualisation obtained is useful when determining the phase of progression of the age series;…”. If the “sustainability data” was analysed alone, would results be much simpler and clearer? Thus, the first central question is what would be the benefit of using HUMP + KPCA + Agglomerative Clustering to determine the phases of progression of the age series? The second central question is how this sustainability data is associated with other attributes in the dataset?

Some minor issues: 

1.     Page 2, Line 65 – 67, repeated sentences.

Author Response

Please, see the attached file with the response to the reviewer's comments.

Author Response File: Author Response.pdf

Reviewer 2 Report

The Authors did not address the Reviewer Comments by properly improving the paper, just expressed their disagreement in the Cover Letter. Hence, the Reviewer’s decision remains the same.

 

Author Response

Please, see the attached file with the response to the reviewer's comments.

Author Response File: Author Response.pdf

Round 3

Reviewer 1 Report

The authors have addressed the concerns mentioned in the previous feedback.