# Decentralized, incentivized research peer reviews

## User Profiling

Uses orcid researcher profiles as foundation for decentralized, web3 provable profiles

building on top of what retrox has achieved
https://github.com/retrox-pgf-team/orcidauth/tree/arxiv-api

https://orcidauth.vercel.app/register

("login with orcid")

## web3 peer review processing

roughly relating to the ANTS peer review grants proposal (EthTurin2020)
https://linktr.ee/antsreview
https://arxiv.org/pdf/2101.09378.pdf

- peer reviews are requested by authors (Issuers in AntsReview)
- anchored on chain, mentioning the paper under review with document identifier (e.g CrunchDAO, or IPFS link), with small token reward for successful review
- anyone can further contribute to the potential reward pool that will be released to a successful review
- issuers will nominate an Approver who will eventually accept/reject reviews
- Approvers need to fulfil a certain minimal condition s.t. they have submitted at least X successful reviews in the past
- issuer and Approver identities are revealed at the end of the peer review process so issuers have a strong incentive to nominate a legit / non collusive Approver
- reviews can be backed by fundraisers or hypercerts (or individual contributors (ANTS))
- goal is: who submits a relevant peer review will get (a part of) the staked incentive

- peer reviewer requirements
- ideally we would authenticate peer reviewers academic credentials automatically from ORCID and whitelist them for review, but a manual step in letting the approver double check the review against the reviewer credentials will suffice
- peer reviewers must have attestation/verified credential in certain research field as guaranteed by their public profile and ORCID-ETH link, they can only review if they have linked their ORCID to their address
- they need to stay anonymous to the author
- e.g. submit a hash <public address, undisclosed address>, or ZK proof that they are one of the verified reviewers
- there cannot be collusion between authors and reviewers, this is later verified by the editor.

- peer reviewer submits the review
- peer reviewers must have attestation/verified credential in certain research field as guaranteed by their public profile
- the nondisclosed address is disclosed on chain as the "anonmyous" peer review address
- Don't disclose their public profile address (or they do, depending on if this is desireably public)
- addReview() called and review published to IPFS

- the editor (Approver) checks the peer reviewer identity and assesses the peer reviews
- Approver must be able to see the reviewers public profile, this can be resolved bc the reviewer sends an off-chain message to the approver
- Approver rejects or accepts review
- contract drops reputation on reviewer if review is accepted
- If review is accepted contract releases financial reward to reviewer and also a small reward to acceptor
- review is published on IPFS with the original submission and acceptor comments

- the peer reviewer is able to prove to others that they have been the ones filing the successful peer review by revealing control over the initial commitment

This protocol works both for standard peer review, as well as for proposals requesting funding. For the latter, the approver (as well as possible other third parties) provide a certain amount of funds which shall be distributed other retroactively or proactively to a number of nominated projects in particular research area. The peer reviewers act as expert judges as to where to assign funds. A small share of the distributed funds will then be distributed to the reviewers to reward them for their work. Similarly, the whole protocol works for standard peer review as explained previously, where the approver functions as an editor or senior reviewer. The issuer requests review for a paper and a financial reward is provided, initially by the issuer, but can be topped up by external funders.