An Effective Negotiation Strategy for Quantitative and Qualitative Issues in Multi-Agent Systems

Round 1

Reviewer 1 Report

Thank you for providing me with the opportunity to read “On Effective Hybrid Negotiation in Multi-agent Systems”. I have the following comments:

·         Please revise the title and make it more meaningful and a proper representative of the content of the paper.

·         The abstract doesn’t mention the problem or issue addressed in the paper. Please add it to the abstract.

·         Negotiations in different contexts require different approaches and variables. The authors need to specify which sector they are referring to. Alternatively, the generalization of the negotiation to all sectors must be justified.

·         Group the clusters of references throughout the paper e.g., lines 26-27 ref 4 to 9 can be written as [4-9]. Please correct this throughout the paper

·         Again the problem is not clear from the introduction and the novelty is not elaborated on. The authors need to identify the problem clearly and then explain the novelty of their study in a detailed paragraph. Specifically, the innovation compared to related studies must be highlighted and discussed.

·         Line 67 what is [?] is there something missing here?

·         The related work needs improvement. Specifically, a critical discussion on state of the art is missing here. Please revise

·         The authors need to justify the need for and highlight the innovation of their model. The authors explain that their model is a simplified version of an already existing model. In the presence of a much more sophisticated model, what is the need to have another model? Again the innovations must be presented and discussed.

·         Where are the quantitive issues in line 127 originating from? The authors need to discuss them properly in related works and provide references. Alternatively, serious justifications are needed for these and other following variables.

·         Please add a proper method section to the paper and discuss how the paper is presented, and justify the need for different sections. Also, the techniques and tools used in the study must be elevated on.

·         What is the rationale for using reservation intervals: small (20%), medium (50%), and high (90%)? The relevant justification is missing

·         The paper needs a detailed discussion section. In the discussion section, compare the findings with other relevant studies and provide references to these. The authors must focus on the key improvements in their study compared to the ones they will compare it with. In addition, the innovations should also be highlighted.

·         Add the limitation of the study to the conclusion section

·         What are the practical and managerial implications of this study?

Author Response

File is attached

Author Response File: Author Response.pdf

Reviewer 2 Report

The paper presents the preferences-based negotiation algorithm, taking into account qualitative and quantitative characteristics of the commodity. The algorithm has interesting properties, but its description should be improved.

The mechanism has not been described in sufficient detail. In particular, it has not been clarified (also formally) what the negotiation issue is (lines 93-103).

Lines 109-117: do you suggest that the negotiating agent knows the reservation price of its opponent? What level of knowledge is assumed? Please explain this.

Lines 167-168: Please explain exactly how the issue is divided into favourable and unfavourable groups.

What exactly is iso-curve? What do x and y mean? What is the value of particular iso-curves? What is their relationship with the Pareto front - a term related to, inter alia, multi-criteria decision-making? I also suggest taking a look at the multi-criteria decision-making theory to organize the concepts.

I suggest that the concept of similarity be formulated correctly (formally) - as a metric in the N-dimensional criteria space. Thanks to this, you will not need to enter a computational example.

As the description of algorithm 1 suggests that it consists of two components - I propose to break it into two parts.

Utility rate, agreement rate and product rate are not described at all, please provide the proper description (preferably using formulas). Also, strategies (basic, preference-based, fuzzy) are not described at all.

Line 66: lack of reference

Author Response

File is attached

Author Response File: Author Response.pdf

Reviewer 3 Report

The article is devoted to the study of negotiations between the buyer's agent and the seller's agent on several issues. The study's relevance is justified by the fact that automated negotiations are an effective way of interaction in which agents exchange offers and counter-offers to conclude a contract. A hybrid negotiation method is presented and tested to improve the results of the negotiation process. The sentence formulation is based on fuzzy similarity and preference methods. The preference-based mechanism is used for quantitative questions, and the fuzzy similarity method is used for qualitative questions. The preference-based mechanism considers the opponent's preferences when forming proposals: the agent makes significant concessions on issues that the opponent prefers more. The fuzzy similarity method formulates a proposal that considers the offer of a deal more similar to the one received by the opponent during the last round of negotiations. The experiments consist of two parts. The first part compares the hybrid strategy with the base one. The results show that the hybrid strategy is better in all efficiency measures, namely utility ratio, agreement ratio, and Nash product ratio. In the second part of the experimental work, four sentence generation mechanisms are compared: fundamental, preference-based, fuzzy similarity, and hybrid. In terms of utility rates, agreements, and Nash product rates, hybrid negotiation works better. On the other hand, the preference-based algorithm outperforms the fuzzy similarity method in terms of the Nash product coefficient.

Despite the satisfactory quality of the article, some shortcomings need to be corrected.

1. It is recommended to expand to abstract with numerical results of the study.
2. The aim of the research should be defined.
3. It is recommended to include papers about decision-making in multi-agent systems in the Related work section, e.g. doi: 10.1109/ELIT.2019.8892307
4. The input parameters used for the experimental studies should be described in more detail.
5. It is recommended to include the Discussion section to compare obtained results with other research.
6. The practical novelty of the research should be highlighted.

In summarizing my comments, I recommend that the manuscript is accepted after major revision. 

Author Response

File is attached

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

The paper still needs significant improvements to be fit for publication:

1. The problem is still not clear in the abstract. "reaching an agreement is not a problem." The authors need to clarify the problem addressed in the study.

2. There is a sense of confusion in the paper. The authors insisted that the paper focus on all domains, then specified it to be focused on the buy-sell process of a car, and then in the introduction suggested it focus on business.  This must be clarified throughout the paper. Again, the fact that negotiations in different context require different approaches need to be justified. How exactly is the business negotiation tackled in the current study? This is not clear in the introduction.

3. I am sure MDPI formatting allows for the grouping of references. Please check again.

4. I can not see where exactly the authors discussed the novelty in the introduction. Specifically, the authors were requested to discuss the innovation compared to related studies. In revision, always mention the exact lines or paragraphs where the comment is addressed.  

5. The related works haven't been properly improved. A critical discussion is still missing, as requested in the previous review round.

6. So the authors agree that the existing model is better than their model? Again what is the need for this model if a superior model exists? This hasn't been clarified. 

7. References were requested for the quantitative issues that haven't been added in the revised version.

8. A proper method section has been requested however, the authors pointed out the organization of the paper at the end of the introduction as an alternate to the method section. The reviewer doesn't agree with this and recommends adding a detailed method section to the paper with a flowchart and follow-up discussion on what methods, tools, and techniques for validation have been used in the paper. 

9. The author's explanation of the comment "What is the rationale for using reservation intervals: small (20%), medium (50%), and high (90%)? The relevant justification is missing" doesn't seem relevant, please check and provide a proper explanation with references. 

10. A detailed discussion section is needed in the paper. It becomes weak and lacks critical discussion. Without such a section where the key takeaways are elaborated on.

Author Response

Please, see the attached file.

Author Response File: Author Response.pdf

Reviewer 2 Report

Thank you for considering my comments.

Author Response

Thank you very much.

Reviewer 3 Report

Thanks for the authors for the deep analysis and considering reviewer's comments and recommendations. In my opinion, now the paper can be accepted.

Author Response

Thank you very much.

Round 3

Reviewer 1 Report

Thank you for addressing my comments. Please move the figures from conclusion section to the section prior to it. They should appear next to where these are mentioned. All the best