ACUX Recommender: A Mobile Recommendation System for Multi-Profile Cultural Visitors Based on Visiting Preferences Classification

Round 1

Reviewer 1 Report

The authors propose a recommendation system based on visitors' preferences for possible points of interest. They also conducted a user study to determine the system's attractiveness, dependability, efficiency, and novelty. This is an interesting and well-written paper. Among the highlights of the paper is how clearly the authors identified their contributions in the introduction, strengthened the paper by comparing it with similar work, provided adequate explanations of the methodologies employed, and conducted a user evaluation to evaluate its effectiveness. As it is a user study, I am unsure of its reproducibility.

Author Response

First of all, we would like to thank you for your effort in revising our manuscript. We really appreciate the careful review and constructive suggestions. In what follows, we try to address all the points raised in the review. The manuscript is now substantially improved after making the suggested edits. More specifically, regarding the reproducibility of the paper, we already begin to test the MRS using a narrower case study, utilising a dataset of POIs at the Stavros Niarchos foundation.

Reviewer 2 Report

The paper presents a mobile recommender system focused on recommending personalised cultural point-of-interests (POIs) to visitors based on their visiting preferences. The main feature of the proposal is the presentation of different approaches to elicit the information required to this specific recommendation context. To this end, the research is supported by the ACUX Typology conceived for the harmonisation of the cultural visitor typologies, and that has been proposed by the same authors. The introduced framework is evaluated through a user study.

The manuscript is interesting and easy to follow. However, we think that the authors should cover some issues that have not been properly addressed according to our viewpoint.

-We think that the Related works sections should motivate better the direction of the new research. Considering this set of works previously developed, why is the motivation for the presented research? The specific contribution to the state-of-art should be highlighted.

-We think that the "equations" across the work should be corrected, because in several case they are actually pieces of algorithms.

-In this way, further formalization is needed across Section 3. Specifically, Section 3.2.3 should provide more details of the stages of the recommendation algorithm, and includes some figure to support this explanation.

-Even though the main part of the evaluation is focused on a user study, we suggest to consider the discussion of the goodness of the discussed proposal in relation to previous similar works.

-We suggest to discuss the incorporation of the group recommendation paradigm in the current scenario. The analysis of some content-based proposals such as the recently referred at "Content-based group recommender systems: A general taxonomy and further improvements", can be used here with this aim.

 At this stage we suggest Major Revision

Author Response

First of all, we would like to thank you for your effort in revising our manuscript. We really appreciate the careful review and constructive suggestions. In what follows, we try to address all the points raised in the review. The manuscript is now substantially improved after making the suggested edits. More specifically, regarding your comments:

1. We highlighted the contribution to the state-of-the-art in the last paragraph of the Related work section. Also, the contribution is highlighted in the Introduction section.
2. We make corrections to the equations. The new equations Equation1 and Equation2 are shown in Sections 3.2.1 and 3.2.3.
3. The addition of the recommendation algorithm pseudocode in Table 1 of Section 3.2 is a further formalisation, while in Sections 3.2.1, 3.2.2. And 3.2.3., more details of the stages were provided.
4. The proposed MRS was evaluated in Section 4 in juxtaposition with 468 studies provided by the UEQ data analysis tool.
5. We consider the group recommendation feature as future work in the Conclusions section.

Reviewer 3 Report

In this manuscript, the authors aim to personalize the recommendations for visitors based on their own preferences, called the ACUX Recommender (ACUX-R). To do so, the most critical part of this study is the method to classify the visitors. This motivation seems general and one of the very traditional issues in the recommendation system area. The reviewer positively assesses this manuscript because it is a well-structured presentation, and is simply understandable. Despite this assessment, the manuscript has several major points which should be substantially improved for publication.

1. It could be interesting if the author summarized the contribution of the study in the Introduction section.
2. Notation and equation: It will be better if authors create a table with all notation and use it consistently in the whole manuscript. The author doesn't even number the equations.
3. The authors named section 3.2 as algorithms but they do not show any equations core or any Pseudocode to show how to run these algorithms.
4. The evaluation section is not suitable for the academic manuscript. There are only 20 participants answered the questions. The author even does not show the performance in generating recommendations. The reviewer suggests authors revise the evaluation part, which provides more details experiment setting, evaluation metrics, and results for the evaluation of the recommender system.
5. The most not good thing about this manuscript is the similarity check gets a very bad result, around 40%. The reviewer suggested the author had to revise the manuscript very carefully before re-submit. I attached the results for reference.
6. In addition, the paper is comprised of many grammar mistakes and unprofessional presentations. I suggest the authors carefully rewrite the paper before re-submission

Comments for author File: Comments.pdf

Author Response

First of all, we would like to thank you for your effort in revising our manuscript. We really appreciate the careful review and constructive suggestions. In what follows, we try to address all the points raised in the review. The manuscript is now substantially improved after making the suggested edits. More specifically, regarding your comments:

1. We highlighted the contribution to the state-of-the-art in the last paragraph of the Related work section. Also, the contribution is highlighted in the Introduction section.
2. At first, we make corrections to the equations. The new equations Equation1 and Equation2 are shown in Sections 3.2.1 and 3.2.3., and also inside the pseudocode table.
3. The addition of the pseudocode in Table 1 of Section 3.2 is a further formalisation to provide information regarding the proposed recommendation algorithm.
4. The evaluation process was an ongoing process. Thus, 50 participants participated in the user study, while even more completed the questionnaire. Also, mean values per scale and confidence intervals per scale as diagrams were added as evaluation metrics.
5. We carefully revised the article, and the examples in Sections 3.2.1 and 3.2.2 provide more information to help the reader comprehend how the algorithm works.
6. The paper was carefully reviewed.

Round 2

Reviewer 2 Report

The authors have covered my previous comments. I suggest paper acceptance.

Reviewer 3 Report

The authors have carefully revised the manuscript. Almost the suggestion from reviewer have been completed. This manuscript is ready for publication.