Multi-Granularity and Multi-Interest Contrast-Enhanced Hypergraph Convolutional Networks for Session Recommendation

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

I had the opportunity to read your paper. I liked the ideas and the way you solved the problem. But I want to have a second read after you improve the paper by incorporating this:

1. Introduction must highlight the problem you are solving and why it's urgent. Base this on examples and literature references.

2. A literature review is mandatory. Highlight the gaps.

3. Add a discussion section and mention applications of your work.

4. Enhance concluding remarks.

Comments on the Quality of English Language

Its fine.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

This article proposes a new multi-granularity and multi-interest contrast-enhanced hypergraph convolutional network model for session recommendation. I have some minor and major comments before accepting this article. 

1. The contribution is not clear. 

2. There are no references to datasets. 

3. How to get the value of Table 2 is not described. 

4. The article is so plagiarized. Authors need to reduce plagirarism. 

5. What is the main objective of Section 3.5 (Ablation Experiment)? 

6. Add more relevant references.

Comments on the Quality of English Language

Needs to rewrite the article in scientific and familiar words.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 3 Report

Comments and Suggestions for Authors

State of the art must be extended and improved.

A paragraph about the rest of the paper must be introduced at the end of the introduction.

Justify metrics. At the same time, expose the expressions and variables of these metrics.

Justify parameters of section 3.3.

Add future work with details and relevance.

Add abreviation section.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 4 Report

Comments and Suggestions for Authors

Define all acronhyms.

Define all variables, constants, symbols and units.

"Deep 30 learning based models mainly include RNN and its variants [4,8-11]" the ANN's require a learnimg model they cannot learninh itself, mention the variants and the learningg models.

Fix the tipo, gramatical and punctuation errors.

Bibliography must be in order. Reference 18 is mentioned previous reference 16 and reference 17 is mentioned later than reference 18 in introduction. 

References 1-3, 5-7 are missing.

The introduction needs to be enhanced with bigger analisys of the literature.

Formulation of the problem in first paragraph is confuse needs to be enhanced and recommend placed it as formulas referenced and explained in text.

Fig. 1 is deformed and so blurry needs to be enhanced. needs bigger explanation.

Define every element in every equation to better understanding of the reader. 

There is no relationship between the some paragraphs and the mention of variables and parameters mentioned above or simply not mentioned makes the paper unreadable.

There are no definition of the mechanism used to generate the change from linguistic to numbers with limis, ranges, among others.

How determinate the parametrization.

Why this values are selected.

What means the difference on performacne values mentioned in tables 2 and 3. Explain widely.

Fig. 3 is to blurry needs to be enhanced.

All bibliography proced form congress why not appear paper or other sources of information on the state of art.

 

 

 

 

 

Comments on the Quality of English Language

there are gramatical, punctuation and tipo errors.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 5 Report

Comments and Suggestions for Authors

The paper proposes a novel Multi-Granularity and Multi-Interest Contrast Enhanced Hypergraph Convolutional Network (MGMI-HCN) architecture for session recommendation. The approach is both theoretically sound and practically appealing. By effectively incorporating multi-granularity and multi-interest contrast, MGMI-HCN demonstrates a superior ability to capture complex user preferences and item relationships. This well-described and thoughtfully constructed model offers a promising avenue for advancing the state-of-the-art in session-based recommendation systems.

- To see the projects, categories and prices as a graph is a good idea.

- For M-G Encoder, lines 120 to 123.  Is there any information theory metric that you computed to say that  (Obviously, nodes between the same granularity contain the same semantic information, while nodes between different granularities contain different semantic information)

The papier is well structered, well produced and deserve acceptance.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

All my comments are addressed. Congratulations.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

This present form can be accepted. 

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 4 Report

Comments and Suggestions for Authors

I recommend add more fererences to papers published in journals, books, book chapters.

Author Response

Please see the attachment.

Author Response File: Author Response.docx