Matching Consignees/Shippers Recommendation System in Courier Service Using Data Analytics
Round 1
Reviewer 1 Report
The paper provides an unique combination of data analysis techniques to address a specific set of questions related to a case example in a particular location. The selection of the techniques is well justified from theory. Analysis is then conducted competently and described well, using a specific dataset. The results from this analysis are specific to the case and data examined. However, the combination of techniques is valid for other similar cases. I believe the paper is worthy of publication.
Author Response
Dear Reviewer.
Please find the attached answers to the reviewer points by points of the revised manuscript entitled “Matching consignees/shippers recommendation system in courier service using data analytics” (Manuscript ID: applsci-851751). The manuscript has been revised according to the reviewers’ comments for your consideration to be published as a research article in Applied Sciences. If you have any more questions, please let us know.
Looking forward to hear from you very soon,
Best regards,
Authors
Author Response File: Author Response.pdf
Reviewer 2 Report
I suggest deeply research about previous work on this topic.
Author Response
Dear Reviewer.
Please find the attached answers to the reviewer points by points of the revised manuscript entitled “Matching consignees/shippers recommendation system in courier service using data analytics” (Manuscript ID: applsci-851751). The manuscript has been revised according to the reviewers’ comments for your consideration to be published as a research article in Applied Sciences. If you have any more questions, please let us know.
Looking forward to hear from you very soon,
Best regards,
Authors
Author Response File: Author Response.pdf
Reviewer 3 Report
Summary
The manuscript aims to develop techniques and strategies in order to provide recommendation to consignees/shippers in courier activities. In order to do so, they use data analytics methodologies, such as clustering techniques and decision trees, while Monte Carlo simulation is used in order to evaluate the impact of their recommendation on the profit of the system. Through computational experiments, the authors show the effectiveness of their strategy.
Referee’s comments
The manuscript is interesting and appropriate for this journal. However, there are some concerns. First, the authors should revise their manuscript from the language point of view because some sentences are not clear. Second, the literature review should be enriched and made complete. Third, more details and motivations should be provided in order to support the choice of methods.
Consequently, my recommendation is for a major revision.
Specific comments
- In the abstract, I would suggest to avoid the expression “the most optimal”.
- At line 87, at page 2: the authors should correct the repetition “identify develop”.
- Line 103 at page 3 is not clear.
- I suggest the authors to have a look at the paper “Matching supply and demand in a sharing economy: Classification, computational complexity, and application” by Boysen et al. and, if appropriate, to add it to the literature review.
- In the literature review, the authors should not only explain what is done in other papers, but also compare those papers to their manuscript.
- In the methodology section, the authors should provide more details concerning the pre-processing phase of the data. For example, they should provide details on the imputation techniques used in order to obtain missing data.
- At page 7, the authors should motivate why they used the k-means clustering algorithm. In fact, it is known that the k-means clustering algorithm requires that the number of clusters is pre-specified at priori. For example, the authors could compare their results when using the k-means clustering algorithm to the ones obtained by using a hierarchical clustering method, which does not require to specify the number of clusters at priori.
- Does the confidence in Table 3 represent the p-values used when testing some hypothesis? If yes, then the cutoff value should be for p-values lower or equal than 0.005 (or 0.001).
Author Response
Dear Reviewer.
Please find the attached answers to the reviewer points by points of the revised manuscript entitled “Matching consignees/shippers recommendation system in courier service using data analytics” (Manuscript ID: applsci-851751). The manuscript has been revised according to the reviewers’ comments for your consideration to be published as a research article in Applied Sciences. If you have any more questions, please let us know.
Looking forward to hear from you very soon,
Best regards,
Authors
Author Response File: Author Response.pdf
Reviewer 4 Report
The authors propose a Recommendation System to match shippers with consignees, by using an association rule, a clustering technique and a decision tree. The methodology is applied to a transportation company in Thailand. Finally, a Monte Carlo simulation allows to estimate the benefits from the resulting assignment.
Due to significant advances in Big Data and the Internet of Things, recommendations systems are becoming increasingly for companies. Therefore, I think this work can be of interests for researchers and experts alike. In addition, the paper seems to be sound.
Author Response
Dear Reviewer.
Please find the attached answers to the reviewer points by points of the revised manuscript entitled “Matching consignees/shippers recommendation system in courier service using data analytics” (Manuscript ID: applsci-851751). The manuscript has been revised according to the reviewers’ comments for your consideration to be published as a research article in Applied Sciences. If you have any more questions, please let us know.
Looking forward to hear from you very soon,
Best regards,
Authors
Author Response File: Author Response.pdf
Round 2
Reviewer 3 Report
The authors have taken into account all my recommendations.