Over the last few months, we’ve developed an implementation on top of our platform on the blockchain backend in which the publishers within this initiative will store peer review data.
We have developed our platform on top of an Ethereum compatible blockchain, called Ethermint. Ethermint is an implementation of Ethereum built on top of Tendermint. At a high level, the blockchain consensus engine, called Tendermint Core, and the generic application interface are used from Tendermint. Due to the flexibility in the way the application interface (Application BlockChain Interface (ABCI)) is designed and developed, transactions can be processed in any given programming language. Ethermint uses Ethereum Virtual Machines (EVM), meaning that it has the full capability of writing and executing smart contracts in Ethereum. Keep in mind, the Tendermint Core (i.e. the consensus engine) is Proof-of-Stake, meaning we’re not abusing the environment and consuming huge amounts of energy as Proof-of-Work consensus algorithms, a question we are often asked.
Proof of Concept
Our Proof of Concept takes data from editorial systems and we extract some information from it, mostly around a manuscript and a review.
We split this data into two streams:
- One stream populates a reviewer’s ORCID profile. This data is Personal Identifiable Data (PID) and falls under the General Data Protection Regulation (GDPR).
- The other stream stores obscured data and relationships between data entities describing a review event on the blockchain backend.
The data stored in the blockchain retains the full anonymity of the reviewer’s identity while allowing a reviewer to prove algorithmically that they have performed the review. We store the states of the review and the connection between the manuscript and the anonymised reviewer. When we store data on the blockchain it is only data agreed by the publisher to be public information and unidentifiable information for any person (reviewer) involved.
In ORCID we store enough information to show that a reviewer has performed a review in a given time frame for a given journal. We store a way to go back to the review itself and to the record in the blockchain but that is obfuscated and protected through a permission layer. This implies that if you don’t have the key to decrypt or the required permissions you cannot go back to the record on the blockchain and the information is meaningless.
The next steps will be to seed the system with data from more journals, allowing us to further demonstrate data querying capabilities across multiple journals, or even multiple publisher sources. We will share some of these outcomes in a few months.