Improving the Precision of Image Search Engines with the Psychological Intention Diagram

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The paper presents an approach for enriching image search engines with Psychological Intention Diagram that models user intent during search. The idea seems interesting, however, the paper needs some improvements.

First, the motivation of using de PID is not sufficiently motivated and discussed in relation with the extensive work in search engines. The relation of search intention with search context needs to be clearly discussed as the examples provided can be seem simply as a problem of context-aware search. Second, the use of PID needs to be properly justified. For example, why PID and no other approaches? Also related works section needs to clarify the advantages of the proposed approach in regards with other state-of-the-art ones.

The other weal point of the paper is the experimental evaluation. The methodology in general is not well described and the results are only evaluated on a set of simple problems. A more detailed description of the methodology needs to be done in the paper, including details about how the metrics were obtained, the agreements among users, etc. More importantly, the results are not compared with any baseline, this is crucial to determine the validity of the approach.

 

Comments on the Quality of English Language

English writing is understandable, minor corrections are required.

Author Response

Thanks for your precious comments. The reply as well as the revision of the manuscript is made. We cherish the opportunity. Hope we can a chance to publish the paper in this prestigeous journal. 

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

 

1. Novelty and Introduction:

   - The introduction should be revised to clearly highlight the novelty of your work. Explain how it advances the field or presents a new perspective.

   - In the discussion of novelty, provide a more detailed and focused explanation of what sets your research apart from existing work.

 

2. Language and Clarity:

   - The English writing in the manuscript requires improvement. There are instances where words and sentences are confusing or unclear. Consider proofreading and, if necessary, seeking assistance from a language editor.

 

3. Algorithm Figures and Diagrams:

   - The figures and diagrams representing your algorithm need to be more polished and clear. They should effectively convey the key concepts and steps of your algorithm.

 

4. Experiment Design and Implementation:

   - The design and implementation of your experiments need to be better explained and organized. Readers should have a clear understanding of your experimental setup, variables, and procedures.

 

5. Choice of Method:

   - Justify your choice of using GMM over more contemporary, state-of-the-art (SOTA) methods. Explain the rationale behind this decision and discuss how it impacts the results.

 

6. Experiment Scaling:

   - Consider expanding your experiments to a larger scale dataset. This will help assess the generalizability and robustness of your approach under different conditions.

 

7. Additional Metrics:

   - Include a more comprehensive set of metrics in your experiments to provide a thorough evaluation of your approach. This will enhance the credibility of your results.

 

8. Overall Quality:

   - The current version of the manuscript falls short of the journal's standard. Revise and address the issues mentioned above to ensure that the paper meets the necessary quality criteria for publication.

Comments on the Quality of English Language

See above my commnets 

Author Response

Thanks for your precious comments. The reply as well as the manuscripts have been revised according to the suggestions. 

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors This paper utilize the psychological intentions of the web page designers to construct the PIDD and 731 the psychological intentions of past users to construct the PIDU. My comments are as follows: 1. This can also be achieved using text-driven retrieval methods in the following documents, why don’t the authors compare them? Or can be developed as part of related work.         Exploring a fine-grained multiscale method for cross-modal remote sensing image retrieval         Remote sensing cross-modal text-image retrieval based on global and local information         Hypersphere-based Remote Sensing Cross-Modal Text-Image Retrieval via Curriculum Learning         MCRN: A Multi-source Cross-modal Retrieval Network for remote sensing 2. The algorithm part does not have clear input and output, please further improve it. 3. For retrieval tasks, why not use R@k for evaluation? Can you give relevant comments similar to the following articles?         Stacked Cross Attention for Image-Text Matching         A lightweight multi-scale crossmodal text-image retrieval method in remote sensing 4. The full text does not provide formal modeling of the method. It is recommended that the author add some formulas to facilitate readers to better understand. Comments on the Quality of English Language

ave

Author Response

Thanks for the precious comments. The revision is revised according to the comments as possible. 

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

 

Although the authors try to answer to my original concerns in the response letter, the changes in the paper itself are minor and do not include these explanations. For example, the response mention some addition on the beginning of section 3 that is not present in such section. I still think that the differences with other approaches, for example, need to be more clearly stated. Also my concerns about the validation of the approach using a few examples remains and the response is not convincing.

Comments on the Quality of English Language

English writing is readable.

Author Response

     We appreciate the precious comment. We do our best to revise the paper to meet the requirement of reviewers and have highlighted the revision part in red for this and last revision, hoping the revision can meet the desired.  

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

No more comments

Comments on the Quality of English Language

well

Author Response

Comments:  No more comments

Reply: Thanks for the precious comments. We sincerely thanks for your suggestion to enhance the completeness of the paper. 

Round 3

Reviewer 1 Report

Comments and Suggestions for Authors

 

Most of the concerns from my previous reviews were addressed in the revised version of the paper. I still think the main problem in on the validation of the approach, as experiments are not sufficient (in number) for draw conclusions. Regarding presentation, the paper has some sections that are too long. For example, the introduction takes more than two pages, it should be more concise and state the contributions more clearly.

Comments on the Quality of English Language

There are some mistakes in the text (e.g. varaibles for variables), but the paper is in general readable.

Author Response

We appreciate the comments, which enrich the contents, emphasize the contribution, and increase the readability of this paper. We hope we can share our research results in this prestigious journal with the research community. 

Author Response File: Author Response.docx