“Follower of the Reference Point”: Platform Utility-Oriented Incentive Mechanism in Crowdsensing

Round 1

Reviewer 1 Report

## Summary

The authors propose an advancement for crowdsensing platforms using monetary incentive mechanisms. The proposed incentives are based on prior experience founded on theories of behavioural economics. The authors claim that their algorithms enhances the task completion rate and chooses more potential participants.

## Critique of the proposed idea

I read the paper and found it very interesting. 

However, there are some points which need to be improved. 

Please consider all of the following points:

- In the abstract and introduction please motivate your research by one or two concrete use cases. There are crowdsensing use cases where there is no bidding involved. There are also many crowdsensing applications where a participant does not have to make any decisions. This confuses the user. 

- Explain in detail which kind of risks the participants face. There are risks of not earning money, risk of loosing money, risks of overwhelming participants, health risks, … . Please be concrete.

- Research questions, that drive the paper, should be built in the introduction from an ongoing and pertinent bibliography. Identifying a research gap is not enough; key is showing its significance to the field.

- Detail your research design and research method.

- You distinguish between new and old participants. What role does the multiple repetition of a participation play? How do you recognize this feature in your platform or in your validation? If you store this information, please explicitly describe how data privacy aspects are tackled. 

- Figure 3: between step 7 and 8 the whole completion of the tasks are missing. Is this information used in a way? 

- 304: How does a participant or you determine Cji? Do you have evidence that a participant really knows these costs and/or takes them into account?

- Chapter 4: The text is overcomplicated and difficult to understand. Please summarise your findings and results in easy words. How are the incentives between new and old participants distributed? What are the effects of lower OR or IR?

- 391: Where do they get the risk values from and how did they test these assumptions in their experiments? 

- Algorithm1: Explain the input values. e.g. RUj is never used up to this point.

- Chapter 5: This chapter needs to be completely revised. 

Please explain in detail the framework and settings of your experiments. Is this actual or simulated data? Where do they come from? If it is simulated data, how was its validity checked? If they were real experiments, please detail what tasks the participants had to perform and what the data were. Provide a link where the recorded data can be viewed and comprehended. How were returning participants identified and assessed?

If it is a pure simulation, the results have be backed up with a real experiment.

Minor changes:

- 32 - please explain PM2.5

- 113 - please explain VCG

- 489 - please explain DSTA and how is works

I look forward to receiving your revised manuscript.

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 2 Report

The paper starts well. I liked the first two sections. It connects to previous research, establishes its niche, and states its contributions. Afterward, it becomes a bit hard to follow. For me, it wasn’t that clear how the simulations were conducted. Who were the participants, the tasks they had to complete, and the rewards?

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

In this paper, a platform utility-oriented incentive mechanism in crowdsensing is presented.

The analysis is clear and the results are promising. Anyway, some improvements are required:

1. Fig. 2 should be better described. It is only mentioned as a physical model of IMBE.

2. Fig. 3 is very useful but it is difficult to read. Please increase the size and the quality.

3. How do the parameters in Table 2 have been selected?

4. The conclusions section should address the possible future development of this research.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Thank you for the updated version.

Reviewer 3 Report

All the requested changes have been performed and the quality of the paper has been improved.