Dynamic Graph Convolutional Network-Based Prediction of the Urban Grid-Level Taxi Demand–Supply Imbalance Using GPS Trajectories
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsSubstantive comments:
1. There is no separate introduction section, which should describe the main ideas and problems of the urban taxi services. There should be a separate Literature review section, which should describe the previous research in the field of using several techniques for urban taxi services. Please underline the major novelties in the introduction.
2. Although the authors have provided some literature, but American and European literature is missing (solution to similar problem).
3. Lack of explanation of all symbols used in mathematical formulae in one place.
4. Line 385: Data from January 2019 is not very recent. It should be based on more recent data.
5. Line 401-419: Text extracts from the publication: Yang, H.; Zhang, X.; Li, Z.; Cui, J. Region-Level Traffic Prediction Based on Temporal Multi-Spatial Dependence Graph Convolutional Network from GPS Data. Remote Sensing 2022, 14, 303.
6. Section 5: The discussion should not only be a summary of the results. In the "Discussion" section to compare your own research with the analyzes of other authors, this will additionally increase the value of the publication by starting a scientific discussion with other researchers.
7. Section 6: Conlusions - there is no clear indication of the limitations of the proposed approach.
Editorial errors:
1. Sections should be numbered from 1.
2. Figure 1 – description of the figure: (a), (b), (c), (d); see instructions for authors (IJGI) https://www.mdpi.com/journal/ijgi/instructions
3. Figure 2 – description of the figure “Illustration”
4. Figure 3 – description of the figure: (a), (b), (c), (d), (e), (f), (g), (h); see instructions for authors (IJGI) https://www.mdpi.com/journal/ijgi/instructions
5. Figure 4 – description of the figure: (a), (b), (c); see instructions for authors (IJGI) https://www.mdpi.com/journal/ijgi/instructions
6. Figure 6 – description of the figure - the description should refer to the drawing shown left (a) and right (b).
7. Figure 7 – description of the figure: (a), (b), (c), (d), (e), (f); see instructions for authors (IJGI) https://www.mdpi.com/journal/ijgi/instructions
8. No spaces in front of the bibliography numbers in many places (see annex).
Comments for author File: 
Comments.pdf
Author Response
We gratefully thanks for the precious time the reviewer spent making constructive remarks.
We have made changes addressing the issues you raised. Please refer to the attachment we have submitted for details.
Author Response File: 
Author Response.docx
Reviewer 2 Report
Comments and Suggestions for Authorsthe paper is quite clearly written. In future research, it is necessary to determine whether there are changes in traffic congestion, i.e. whether there are enough locations / drop-off points for collecting passengers in the observed area. For the aforementioned, a simulation of traffic flows should be made
Author Response
We gratefully thanks for the precious time the reviewer spent making constructive remarks.
We have made changes addressing the issues you raised. Please refer to the attachment we have submitted for details.
Author Response File: Author Response.docx
Reviewer 3 Report
Comments and Suggestions for AuthorsSummary/Contribution: The paper proposes a model that combines dynamic graph convolutional networks (GCN) and gated recurrent units (GRU) to predict the imbalance between taxi supply and demand at the grid level in urban areas, using taxi trajectory data. The model captures both the dynamic inter-grid influences in the spatial dimension and the trend and periodic changes in the temporal dimension, resulting in superior performance compared to classical time-series models and spatial-temporal GCN models. Comments/Suggestions: 1. Include a detailed explanation of the methodology used for data collection and preprocessing to ensure transparency and reproducibility of the results. 2. Provide a clear definition and explanation of the grid-level imbalance index used in the study to help readers understand the evaluation metric. 3. Conduct a comparative analysis with other state-of-the-art prediction models to further validate the superiority of the proposed model. 4. Include a discussion on the limitations of the proposed model, such as its applicability to different cities or the impact of varying traffic conditions. 5. Consider incorporating additional features or data sources, such as weather conditions or events happening in the city, to enhance the predictive accuracy of the model. 6. Provide insights into the practical implications of the research findings, such as how taxi companies can utilize the predictions to optimize their dispatching strategies. 7. Conduct sensitivity analysis to evaluate the robustness of the model by varying parameters or assumptions. Consider exploring the potential for real-time prediction and integration with existing taxi dispatching systems for immediate application. 8. Consider exploring the potential for real-time prediction and integration with existing taxi dispatching systems for immediate application. 9. The authors are invited to include some recent references, especially those related to Deep Convolutional Neural Networks. 10. For instance, the authors may include the following interesting references (and others): a. https://www.mdpi.com/2073-431X/12/8/151 b. https://www.taylorfrancis.com/chapters/edit/10.1201/9781003393030-10/learning-modeling-technique-convolution-neural-networks-online-education-fahad-alahmari-arshi-naim-hamed-alqa
Comments on the Quality of English LanguageCan be improved.
Author Response
We gratefully thanks for the precious time the reviewer spent making constructive remarks.
We have made changes addressing the issues you raised. Please refer to the attachment we have submitted for details.
Author Response File: Author Response.docx
Round 2
Reviewer 1 Report
Comments and Suggestions for Authors Dear Authors,Thank you for your answers. My comments are on the editorial side, as not all have been taken into account.
1. Equations 18 and 19 – what is the mathematical symbol? - see appendix
2. Throughout the text, correct the spelling of the references in the text as follows (line 33) “… demand [1].” (no “… demand[1].” 3. Distances elements of Figure 1 and Figure 3; e.g. Figure 1a-d: - see appendix
Comments for author File: Comments.pdf
Author Response
We gratefully thanks for the precious time the reviewer spent making constructive remarks.
We have made changes addressing the issues you raised. Please refer to the attachment we have submitted for details.
Author Response File: Author Response.docx
Reviewer 3 Report
Comments and Suggestions for AuthorsThe authors considered my comments and suggestions
Comments on the Quality of English LanguageA final proofread would be useful
Author Response
We gratefully thanks for the precious time the reviewer spent making constructive remarks.
We have made changes addressing the issues you raised. Please refer to the attachment we have submitted for details.
Author Response File: Author Response.docx
