The Identification of Ship Trajectories Using Multi-Attribute Compression and Similarity Metrics

Round 1

Reviewer 1 Report

Dear authors

The paper presents the classification of the ship trajectory based on the trajectory similarity matrix with multiple attributes (four complex attributes). The experiment was made in the waters of the Bohai Sea and is based on real AIS data. The authors used the applied algorithm to group the trajectories of group ships and separate the distribution of traffic in the waters.

However, this work can be found on the page in the same form https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4545122 - per print version (from august 2023) . Please explain?

My recommendation is that Fig. 2 should be on page 5 to make it easier for readers to follow the individual steps, so only after the description of Fig. 2 and the individual steps should you move on to picture three and its description. That is, before chapter 2.1.2, place Fig. 2 where it belongs according to the content itself. A similar concept should be the rest of the work.

It is necessary to match the number in the picture with the text, for example in line 568, 569, 578,579... etc. it does not refer to Fig. 1 thigh Fig. 15. The same is true with line 603 onwards - it does not refer to Fig. 2.

Is it necessary for one row in the table to be in order, if so why (Table 5)?

Line 608 – Image 26???? that's not a good number of Fig. It is also necessary to make corrections in the text.

Line 623 Image 3? - please correct it so that the pictures follow the text and the order is correct

Line 636 – image number label?

I ask the authors to match the number of images in the entire paper when referring to the image number! The connection must be clear.

I ask the authors to update the list of References with more recent literature (2022, 2023).

English is good, it's just necessary to double-check everything (small mistakes).

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

This paper proposes a method to analyse maritime traffic based on various techniques and metrics.

It is a good paper, which proves its point, clearly written. In my opinion, it is fit for publication, provided some (rather minor) elements be addressed.

In detail:

In paragraph (3) and (4), lines 103-128, you could add some additional literature. There is research that has been done on the topic of vessel to route association (doi 10.1016/j.oceaneng.2023.114755), vessel trajectory clustering (doi 10.1109/ISPA-BDCloud-SocialCom-SustainCom57177.2022.00080) and the use of DBSCAN in the maritime domain (doi 10.3390/e15062218)

Line 152, ‘repair’ is a poor choice of vocabulary. You can use ‘approximate’ instead

Line 171, Traj equation is not quite clear. Please reformulate

Line 206-208. Not clear, please rephrase. Also, 3769 data => 3769 data points.

Line 223, in literature, 0.5 knots is a common threshold between stopped vessels and manoeuvring vessels. See (doi 10.3390/jmse10122012). Can you back this unusual choice of 1 kn by some literature on the topic?

Line 232, ‘repair’ is used again

Figure 5 (b) and (d): it is unfortunate to use such representation technique to represent circular values. The big lines showing that the cog is going back and forth is misleading, as it actually doesn’t move much

Lines 351-352 please express speeds in kn

Figure 10 is very heavily charged with useless information, and is hard to read

Line 397, beware, time is not a native feature of AIS. Timestamps are usually added by the receiver antenna upon parsing

Table 3: number of tracks, or number of data points?

Figure 14: are those trajectory demarcation lines official TSSs?

From section 3.3 on, there is a MAJOR issue with figure numbers and quotes. Please correct all the document and double check that all is good

Line 624: ‘this side’ => which side?

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

The original achievement of the authors of the paper is the synthesis of a method and algorithm for classifying ship trajectories, using modified multi-attribute compression and similarity metrics.

Comments:

1. To better reflect the content of the paper, I propose the following title: "Identification of ship trajectories using multi-attribute compression and similarity metrics."

2. In line 139, the definition of SCCH score is missing, which appears only in lines 473-482.

3. Conclusions, mostly contains the summary of the paper, specific quantitative and qualitative research results should be included here.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Dear authors

Thank you for the changes you made.

 Minor editing of English language required.

Author Response

Good luck with your work, and I hope the article will be published soon.