DeSci-21cm, a decentralized platform for peer-reviewing and publishing research papers

Published in

Coinmonks

30 min readApr 24, 2023

Abstract

DeSci-21cm is a decentralized platform for publishing research papers and peer-reviewing. By using blockchain technology and IPFS distributed storage, we hope to transfer the process of peer-reviewing and paper publishing to be managed by smart contracts. This allows academic publishing to be free from the control of traditional academic publishers and to return to the autonomy of the academic community, with an automatic redistribution of economic benefits throughout the entire chain. This paper introduces the design and implementation of the DeSci-21cm platform, including the creation of journals, paper submissions, reviews, and publishing processes. It also discusses the considerations and principles behind product design and technical implementation. At the same time, we also explore the future plans of the platform and possible issues.

Introduction — Status quo and Motivation

Publishing academic papers in journals is the most important way for the academic community to openly present research results. The academic community judges whether a paper should be published through a process called peer review. Since the 1970s, the market share of the top five academic publishers (Reed-Elsevier, Taylor & Francis, Wiley-Blackwell, Springer-Nature, and Sage) has been rapidly increasing. Traditional academic publishers use a profit-oriented operating model: scientists receive funding from the government to conduct research, and write academic papers that are provided to publishers for free. If there are color pictures in the paper, a considerable APC (article processing charge) must be paid. The most difficult and important peer review process to judge the effectiveness of a paper is done by reviewers as volunteer work. After the paper is published, readers (mostly researchers in the academic community) need to pay to read it. Universities and libraries often purchase journals under the coercion of the monopoly position of academic publishers. For example, the Peking University Library spent 610,000 US dollars to purchase the Elsevier Science Journal Electronic Database in 2010 [1]. For open access journals, the cost will be transferred to the author, who needs to pay a higher APC. For example, in 2021, the OA APC for Nature reached 9500 US dollars [2]. Therefore, academic publishing can be described as a profitable industry. According to a report in 2019, Reed-Elsevier has a profit margin of up to 40% and a market value of 35 billion US dollars. At the same time, Asian countries account for more than one-third of the world’s total research investment, but there is no large publishing company headquartered in Asia or with an Asian background that can have international influence, can independently move towards the global market, and obtain academic publishing profits. For example, as one of the countries with the most papers published in international journals, China has not yet appeared as a publishing house with international influence, independent access to the global market, and access to academic publishing profits. The large amount of scientific research funds spent by the government every year has become business profits for European and American publishing giants.

The academic community has been vocal about this situation. On April 17, 2023, local time in the United States, Stephen Smith, the editor-in-chief of the top journal NeuroImage in the field of neuroscience imaging, announced on Twitter that all editors of NeuroImage and its sister publication NeuroImage: Reports have resigned because they were dissatisfied with Elsevier’s high APC, and they will establish a new nonprofit open access journal called Imaging Neuroscience [3,4]. In 2021, retired professor of economic history, Zhao Dexin, sued CNKI, because CNKI collected more than 100 of his papers and not only did not pay for his intellectual achievements, he also had to pay to download his own papers. This is actually because after a paper is published in an academic journal, the copyright of the work belongs to the academic publisher, not the author himself.

An influential competitor of academic publishers is the preprint repository. Academic researchers can directly publish their papers on preprint repositories such as arXiv, and interested readers can download and read them for free. This has become a common mode of work and academic communication in fields such as mathematics, theoretical physics, and computer science. However, preprint repositories like arXiv lack peer-review, and therefore the scientific validity of the papers cannot be guaranteed.

Reviewing the process of academic paper publishing, we find that academic publishers in a monopoly position are not the core link. Publishers hand over the papers submitted by academic members to peer reviewers who are also academic members. This indicates that the entire peer-review process can be autonomously handled by the academic community without the involvement of publishers. The emergence of decentralized technologies such as blockchain and InterPlanetary File System (IPFS) provides technical means for this academic autonomy, and decentralized autonomous organizations (DAO) provide a reference organizational paradigm for academic autonomy. Starting in 2022, a movement called Decentralized Science (DeSci) has quietly emerged, and more and more people are thinking about the necessity and feasibility of using decentralized technologies to reshape the academic publishing process [5,6,7].

In some mainstream views, decentralized technologies such as blockchain and IPFS are considered the infrastructure of the third generation of the Internet (Web3). On the traditional second-generation Internet, users’ personal data is controlled by Internet giants, just as the copyright of papers does not belong to authors but to academic publishing giants. The third-generation Internet is considered an “internet of value”, and decentralized community autonomy, value empowerment, and circulation are considered its main features.

The two authors of this article began considering the feasibility of using decentralized technologies to provide new channels for academic members to publish papers in early 2022. After discussing with many scholars in the scientific research community, they decided to solidify the traditional peer-review and publishing process of academic journals into smart contracts running on the blockchain, and use decentralized storage technologies such as IPFS to preserve the intellectual achievements of the academic community. At the same time, they developed a frontend user interface that is easy for ordinary users to use, ultimately forming the “DeSci-21cm” decentralized peer-review and publishing platform. We hope that decentralized technologies can help the academic community return to autonomy and no longer be influenced by monopolistic giants.

Blockchain is a distributed database technology that can be used to record transactions and data and ensure that the data cannot be tampered with. Because of its decentralized nature, no single entity controls it, but it is maintained and verified by multiple nodes in the network. Each node has the same copy, ensuring the security and reliability of the data. The core feature of blockchain technology is the use of cryptography to package each new transaction and data into a block, and then connect the block to the previous block to form a chain, hence the name “blockchain”. Each block contains some transactions and metadata, such as timestamps and transaction hash values. Each block also contains the hash value of the previous block, making the data on the blockchain tamper-proof [11,12].

Smart contracts are computer programs that can automatically execute transactions under specific conditions. They typically run on blockchain networks and can execute transactions and protocols without the intervention of third parties, making transactions more secure, fast, convenient, and transparent. Smart contracts are built on programming languages and can define various conditions, rules, and logic. Once specific conditions are met, they automatically trigger the execution of transactions. Smart contracts can be applied to various scenarios, such as cryptocurrency transactions, voting, credit assessment, insurance, real estate transactions, and more. Through smart contracts, people can achieve decentralized applications, bringing more convenience and innovation to society [10,13].

IPFS is a peer-to-peer distributed file system designed to create a reliable, efficient, and distributed internet file transfer protocol. Its goal is to address some of the problems of the existing internet, such as centralization, reliability, and efficiency. IPFS is based on a technology called “content addressing,” which allows users to find and access files through the hash value of the content, rather than relying on traditional location-based addressing. This makes IPFS more reliable and available and allows users to move files from centralized servers to distributed networks. In addition, IPFS also supports file version control, data integrity verification, offline access, and caching, which can bring more innovation and application scenarios to the internet [8]. Each file or folder in IPFS will have a unique content ID (CID) calculated through the Merkel DAG, and if two files have the exact same content, the calculated CID will be the same.

Because the corresponding mathematical principles ensure that the data stored on the blockchain cannot be tampered with (unless in extremely rare cases where an individual can control more than half of the network’s computing power, but this has not occurred in over ten years of blockchain operation), papers published on DeSci-21cm will be permanently licensed, and the copyright will belong to the author, who can dispose of it at will, such as through licensing and transfer of copyright for future profits. When a paper is plagiarized, the immutable block time of the stored paper will prove the author’s original publication time, thus maintaining their intellectual property rights. A common saying in web3 is “code is law,” and once the review rules are written into a smart contract, they will be public, transparent, and cheat-proof, and the review opinions of multiple reviewers will automatically convert into the final decision. The contributions of the reviewers will also be transparently and permanently saved on the blockchain, and excellent review opinions, as well as excellent papers, have the potential to become valuable treasures in the digital world. The distribution of interests throughout the chain will also be publicly and transparently written into the smart contract, and unlike traditional volunteer labor, the work of editors and reviewers will automatically receive economic incentives. Decentralized storage methods like IPFS will make high-value papers more easily accessible to readers. Of course, an important point is that compared to traditional expensive APC, the economic cost of publishing a paper is very low, especially when using L2 sidechains like Polygon.

Product Design Ideas

Our product takes the form of a DApp. DApp stands for Decentralized Application, which is an application built on blockchain technology. Unlike traditional applications, DApps have characteristics such as decentralization, open source, transparency, and security. They typically consist of a front-end interface, smart contracts, and distributed storage, with decentralization being their most prominent feature. They are not controlled by any centralized organization and are difficult to attack or tamper with.

We aim to use the immutable nature of blockchain to permanently certify papers published on our platform. The most natural idea is to directly write the full text of the paper along with the author’s information onto the blockchain. However, storing data on the blockchain incurs economic costs, as the data will be stored in every node of the network. In addition, paper authors may publish their work in different formats, such as PDF, markdown, or Word documents, and may even include videos, code, Jupyter notebooks, and other formats. If stored on the blockchain, correctly parsing the formats would require additional engineering costs. Furthermore, there are limitations to the size of blocks on blockchain. These constraints indicate that directly storing the original paper on the chain is not a good design.

When we store papers using the InterPlanetary File System (IPFS), IPFS returns a fixed-length content ID (CID), which is the hash value of the entire paper content. The same content will get the same CID. We save the CID of the paper on blockchain, which allows for permanent certification while keeping the cost under control.

We use IPFS to store submitted papers. Each user using IPFS becomes a member of the peer network. When a node wants to find a file it is looking for by CID, it starts searching for it from the nearest nodes. In other words, if a high-quality paper gets more widespread distribution and members of the academic community locally store copies of the paper, then that paper will have a universally faster download speed, shorter search time, and is more likely to exist permanently in the IPFS network. Conversely, if a poor-quality paper is ignored, people will not tend to save copies of it, and the search time required to obtain it will be relatively longer. It may even be permanently lost (although its CID will still exist on the blockchain).

An ideal autonomous academic community storage scenario might look like this: each author saves their own papers locally, as well as papers they frequently read; a lab’s server saves papers produced by the lab; universities and libraries form a digital asset alliance for papers, with each university and library maintaining a large node containing as many papers as possible. In this way, the possibility of papers being lost is minimized, and high-quality papers naturally occupy more storage resources and have a better retrieval experience.

If the ideal state mentioned above is difficult to achieve in the short term, in order to ensure that your thesis will not be lost, you can use Filecoin for storage. Filecoin is a decentralized storage network built on IPFS. It utilizes the content addressing technology and distributed storage features provided by IPFS to provide users with a decentralized storage market, enabling storage providers and demanders to store and retrieve data globally. Filecoin incentivizes storage providers to participate in the network and provide high-quality storage services through a unique “proof of space and time” mechanism. In this mechanism, storage providers need to prove to the network that they have stored a certain amount of data over a period of time, and these proofs will be used to calculate their rewards. At the same time, storage demanders can also choose different storage providers for storage to ensure the security and availability of data. Therefore, Filecoin can be considered as a supplement to IPFS, providing it with more powerful storage and retrieval capabilities, and attracting more users and storage providers to participate in the network through incentive mechanisms. In other words, as long as users spend a little money, they can ensure that their thesis will not be lost for a period of time.

However, considering that our DApp mainly targets academic members from different disciplines, most people are not familiar with Web3 products. If using Filecoin, users will need to use two virtual currencies at the same time, one for the gas fee required for various operations on the DApp (such as submitting a thesis, allocating reviewers, submitting review comments, etc.) and the other for paying for Filecoin storage fees, which undoubtedly increases the usage cost for ordinary users.

Considering this, we will maintain an IPFS gateway only for DeSci-21cm in the early stages of the project. This IPFS gateway will be used for the following purposes:

Saving the papers that users submit on the DApp in the initial stages of the project. If we don’t maintain an IPFS gateway, when an author submits their paper, and other readers may not have had time to download the paper to their local IPFS nodes, the author needs to start an IPFS node that can be always connected to the network on their local machine to ensure that their paper can be downloaded by others. But since our gateway nodes always save a copy of the paper submitted on the DApp, the author doesn’t need to start an IPFS node on their local machine to ensure that the readers can download their paper.
When users search for a paper in the IPFS network using the CID (for example, using the default IPFS public gateway https://ipfs.io), if the number of nodes that store the copy of the paper is low, it may take some time to query, but by connecting to our gateway, they can directly search, view or download the paper. The “view paper” button on our DApp points to our gateway address. Of course, users can also use their own IPFS client to view or download the paper.
Due to network reasons, users in mainland China may not be able to access most IPFS public gateways, but our IPFS gateway can ensure that the academic community in mainland China can access the papers published on DeSci-21cm.

When users use the paper upload function on DeSci-21cm, the paper will be uploaded to our node server and added to the IPFS file system. Users can also submit papers to the IPFS network using other methods, but they need to manually copy the paper’s CID to the DeSci-21cm form to submit it to the blockchain for certification.

Figure 1, a highly decentralized and scalable DApp architecture.

A highly decentralized DApp architecture should be as shown in Figure 1 [9]: Core data is stored in the Ethereum database to ensure that it is never lost, data and files are stored in IPFS, event queries are performed using The Graph (a decentralized query protocol that provides reliable data query services for DApps), and Layer 2 (such as Polygon) is used for scalability. Even the front-end files are hosted on IPFS, so the DApp will never be shut down as long as users have the front-end code saved locally and can always be accessed.

However, in the early stages when the user base is limited, this architecture may face some user experience issues. For example, finding the front-end page through IPFS may take longer to load, and The Graph may face payment issues after a certain number of calls. In order to balance decentralization, user experience, and economic costs in the early stages, our Minimum Viable Product (MVP) product will adopt a compromise solution as shown in Figure 2:

The core data is stored in blockchain contracts, such as the CID of the paper, the CID of the review comments, the wallet addresses of reviewers and editors (on web3, usually represented by blockchain wallet addresses, which correspond to the public keys representing user identity, and our DApp uses the most commonly used chrome browser wallet plugin, MetaMask), and the status of the paper (such as waiting for processing, needing modification, already published, etc.).
The peer review process is automatically executed through smart contract code. For example, when more than two reviewers express their approval, the paper automatically enters the publishing state; when there is one approval and one rejection, the editor is required to add more reviewers, etc.
The paper itself is stored on IPFS. Users can submit a folder containing the paper through the frontend interface. The reason for using the folder submission method is that in many scenarios, users may not only want to provide a paper file but also may want to include corresponding images, code, videos, or provide versions in different languages. This is also very useful in future DApp iterations. For example, when the user includes a metadata text file in the folder, the corresponding metadata can be automatically filled in the frontend form, or the images in the folder can be used for frontend carousel, etc.
The smart contract provides ABI (Application Binary Interface, which defines the communication specification between different program modules) interfaces for all important functions, and the frontend page can directly interact with the smart contract through the ABI to complete the core process of peer review.
Ethereum has more nodes than Polygon, so it is more decentralized. However, the cost of executing smart contracts on Ethereum is also very high, and transactions can be congested. Polygon, as an L2 side chain extension of Ethereum, has greater throughput, faster speed, and lower transaction costs, so a large number of DApps run on it. The actual cost of transactions on the blockchain will vary with fluctuations in gas fees and ETH prices. Based on our estimation, the cost of publishing a single DeSci journal contract on Ethereum may be around $100, and the cost of a user’s operation (such as assigning reviewers) may be a few dollars. In order to cover the gas fees for subsequent editing and reviewer operations, authors may need to pay around $10 when submitting a paper, although this is much lower than the thousands of dollars in publication fees, we still hope that users can have lower usage costs in the early stages of the project. Taking into account the cost, experience, and decentralization, we chose to use the Polygon chain in the early stages. The cost on Polygon is less than one percent of Ethereum’s cost. After estimation, the cost of founding a journal can be controlled at around $1, and the contract will require authors to pay around $0.5 for a single paper to cover the potential costs of editing and reviewing, as well as additional incentives for their labor. The contract administrator can adjust this fee rate according to the actual situation after the journal is established. Ethereum and Polygon are fully compatible, and in the future, if needed, our contract code can create a journal directly on Ethereum without modification.
Storing data in smart contracts on the EVM (Ethereum Virtual Machine, which Polygon also uses) incurs high gas fees, so only the most essential data and data required for contract execution, such as paper CIDs, reviewer and editor information, paper status and review decisions, are stored in the contract. Other data is stored in the form of events. Events emitted by the contract are permanently stored in the blockchain log but cannot be queried directly within the contract. The Graph is a commonly used NoSQL service for retrieving data in Web3. The Graph incurs fees beyond a certain number of calls, so we monitor events emitted by the journal contract on our server and store them in a local SQLite database. SQLite is a lightweight relational database that does not require a license or copyright and stores the entire database in a single file. It is ideal for embedded and small applications and is widely used in various mobile and web application development because it is fast, reliable, low-cost, and does not require a separate server. It is generally considered that the bottleneck of SQLite is concurrent writes, but since there is no concurrent writing problem when monitoring blockchain logs, SQLite is a good choice.
Although hosting the front-end pages on IPFS can ensure maximum decentralization and perpetual use of the platform, as it will never be possible to shut down by any entity (unless everyone on the network abandons it), this approach increases engineering costs and reduces user experience. Therefore, our front-end still uses a centralized form, and there is a centralized backend process that listens to events emitted by the journal and provides fast metadata queries through a backend interface to improve user experience. However, this does not mean that DeSci-21cm can be controlled by a single entity. Even if one day the author of this article shuts down the server for various reasons (such as being too poor to renew the server), the journal contract and core data will still exist permanently on the blockchain and will not be lost. In fact, the only impact of losing the backend is the inability to preview certain paper information, such as abstracts and author information, on the front-end interface, but this information still exists in the paper and can still be accessed from the front-end. Even if the front-end service is also lost, the journal on the blockchain still exists, and users can still interact with the journal contract in other ways or restore the DeSci-21cm service themselves.

Each DeSci journal is a smart contract deployed on the blockchain. By specifying the name when deploying the contract, different contracts can be established, such as the comprehensive journal “Future” that we plan to launch in the first issue. There are two types of contracts: DeSci and PrePrint, the former being a decentralized academic journal and the latter being a decentralized arXiv. DeSci-21cm includes three projects: the contract code, frontend, and backend. We plan to open-source the contract code on GitHub from the beginning, and users can use the source code to verify the contract by knowing the address of the journal on the blockchain. This is a manifestation of the transparency of blockchain and smart contracts. As the first release is still in the MVP stage, there is more room for user experience optimization in the frontend and backend, so we have temporarily postponed the direct public release of the frontend and backend, but we will open-source all the code at an appropriate time in the future.

Product Form and Usage

Here, we only introduce the usage of the MVP version of the product. We will make more user experience optimizations based on actual conditions in the future, so the frontend interface of the product may change. We will release dedicated tutorial videos to explain how to use DeSci-21cm, so here we only briefly describe the core processes and the considerations behind the corresponding designs.

The homepage of DeSci-21cm is an index of established journal contracts. If a user wants to establish their own journal, they can contact us to deploy a new journal contract or deploy one using the contract source code on their own.

Any write operation of a Web3 DApp requires a certain amount of gas fee, usually paid in the form of virtual currency. On Polygon, the virtual currency used is MATIC. The gas fee and the corresponding fiat value of the virtual currency fluctuate in real-time. As of the writing of this article, 1 MATIC is approximately equivalent to 1 US dollar.

The use of preprints is very simple, and its main purpose is to allow users to make a deposit. After completing a paper, the author may feel that it is not mature enough and does not want to publish it immediately. However, they can upload the preprint contract for deposit and share their views before publication. This article does not elaborate on the use of preprint contracts, but rather focuses on the use of DeSci journals.

The DeSci journal contract includes four roles: administrator, editor, reviewer, and author. The management relationship between them is only kept at one level, and the management principle is to select trustworthy people but not interfere with specific matters. For example, the administrator can allocate any number of editors to the journal, and the editors can handle the manuscripts to a certain extent, but the administrator cannot handle the manuscripts themselves. The editor can assign reviewers to the manuscript, and can also replace them before the reviewers submit their comments. However, once the reviewers have submitted their comments, the editor cannot modify or withdraw any existing comments. After multiple reviewers form a resolution, the status of the manuscript will automatically switch according to the resolution via a smart contract, and once switched it can not be undone. All the rules of peer review are written in code in the smart contract, where “code is law”.

The user submits the basic information of the paper, including the paper CID and title, on the paper submission page shown in Figure 3. The user selects the folder containing the paper and submits it. The paper will be uploaded to our IPFS gateway server and added to the IPFS network, and the returned CID will automatically be filled in the corresponding form. If the user does not use our page’s file upload function, they can also use other methods, such as uploading the paper or folder to IPFS desktop client manually and copying the CID into the form and submitting it to the contract, and ensuring that the paper copy always exists on the IPFS network so that others can also obtain a copy of the paper. However, we recommend using the page’s file upload function to maintain consistency of the user experience. When the author submits the paper, they need to pay a fee in virtual currency (which will be automatically filled in the form), with a default minimum of 0.6 MATIC. The fee includes two parts: one part is used to cover the gas fee that may be incurred by the editor and reviewer’s subsequent operations, and the other part is the incentive paid to them. Users can manually increase the fee, but can not decrease it below the minimum value. Only the administrator can adjust the specific value of the minimum value.

After the paper is submitted, it will appear on the to-do list of all assigned editors, and any editor can process it. However, once an editor has taken action on the manuscript, such as assigning reviewers or directly rejecting it, other editors cannot handle it unless the administrator replaces the responsible editor. Because changing editors requires additional gas fees, administrators will not do so unless there are special circumstances. In the editor’s manuscript processing page shown in Figure 4, the editor can preview the paper information and choose to assign 2–3 reviewers or directly reject it (in which case the editor can submit review comments). Before the final decision on the paper, the editor can choose to add reviewers, but only up to three reviewers at most, and the final decision requires at least two reviewers (unless the editor rejects it directly). If a reviewer has not taken action for a long time before the final decision is made, the editor can choose to replace them. However, once the editor assigns reviewers, they cannot interfere with the review decision on the paper and the decision-making power is given to the reviewers.

Figure 4, the page for editors to handle manuscripts.

The upper part of the page displays the basic information of the paper, which is consistent with the paper details page that all users can see. In addition to the title and abstract of the paper, users can choose to expand or collapse some information related to the paper. The paper details page will not be described here.

On the homepage of authors, reviewers, and editors, papers related to the corresponding roles and wallet addresses are classified and displayed in collapsible lists. For example, in the homepage of a certain editor shown in Figure 5, the list of papers waiting to be processed is displayed at the top by default. Clicking “more” will take you to the corresponding details page, which shows the specific information of the paper. You can click “view paper” to view the original paper, and different jump links will be displayed according to different role identities, which can jump to review pages, etc.

After the assigned reviewers submit their review decisions on the review page, which has options to modify, accept, or reject the paper, the smart contract automatically updates the paper status based on the decisions of multiple reviewers. For example, if two or more reviewers give a positive review, the paper will directly enter the publication status. If there is at least one reject decision and no more than one accept decision, the editor will be prompted to add more reviewers. Reviewers can also upload a file with their review comments and suggestions, which can be viewed and responded to by the author on the paper details page, as shown in Figure 6.

If two or more reviewers suggest modifications, the author needs to revise and resubmit the paper. The revised paper will have a different CID from the original version due to the content changes. The page for submitting the revised paper is similar to that for submitting a new paper, but with an option to expand and view the relevant information of the previous version at the top of the page. Additionally, the assigned reviewers and editor for the revised paper are automatically inherited from the previous version without manual reassignment. The previous and new versions of the paper are connected by a doubly linked list data structure, and users can view any version of the paper through this structure.

The rules of peer review for papers are reflected in the open and transparent code of the smart contract, as shown in Figure 7.

Figure 7, Flowchart of peer-review code in the journal contract.

Economic System

As a “internet of value”, one of the characteristics of Web3 is that it includes an economic system that considers game theory. Because the operations of the aforementioned editors and reviewers require a certain amount of gas fees, authors need to pay a fee when submitting a paper, a portion of which will be used to cover the gas fees that editors and reviewers may incur. This will be paid to them in a lump sum when they first operate on the paper, according to the default values set in the contract. Since reviewers typically only need to operate on a manuscript once (submitting a revised manuscript is equivalent to submitting a new paper and requires payment of this fee), while editors may need to operate multiple times (e.g., adding reviewers), the default allocation of gas fees to editors will be higher than that to reviewers. We set the default value of gas fees allocated to editors and reviewers currently far higher than our estimate of actual costs, so that reviewers and editors will not need to pay out of their own pockets under normal circumstances.

Another portion of the fee paid by authors when submitting a paper is an incentive, which is not paid to reviewers and editors in a lump sum, but rather distributed proportionally to their weights in each operation, until the incentive is exhausted. Reviewers typically only have one operation, but editors can have multiple operations. To prevent editors from repeatedly operating in order to obtain incentives, we set the default upper limit for the number of times an editor can operate to three. After that, editors can still operate normally, but they will not receive any further incentives.

All users can view the default values for gas fee allocation, incentive allocation weights, and the maximum number of operations an editor can receive incentives for on the admin page. However, only administrators can modify these default values according to actual circumstances. Modifying default values requires administrators to follow certain rules, such as the minimum gas fee paid by the author being greater than or equal to the gas fee allocated to the editor plus three times the gas fee allocated to each editor; otherwise, the contract will be rejected for obvious reasons.

From the perspective of contract security, the gas fees and incentive fees allocated to editors and reviewers are not sent directly to their corresponding wallet addresses, but rather to the account balances of the corresponding wallet addresses in the contract. Editors and reviewers can withdraw their funds on the withdrawal page. This is done to prevent hackers from launching re-entry attacks on the smart contract. There is a default minimum value for withdrawals, as withdrawals themselves require gas fees, so withdrawals with too small an amount will waste resources. Administrators can modify this default value.

It seems that all costs are borne by the author, even though the amount is only one thousandth of the traditional journal’s APC. However, on the DeSci platform, the copyright of the work belongs to the author rather than the platform, so the author can handle the copyright and profit from the work, such as by authorizing it to corresponding science writers. In the future, we may plan a decentralized copyright trading platform to serve the DeSci-21cm project.

Security considerations

As DeSci-21cm is aimed at academic users, scholars outside the decentralized technology field may be unfamiliar with blockchain technology. Therefore, it is recommended to search for and review the basics of blockchain technology and its corresponding security issues before using it. Here are the top three security considerations:

Do not set your own private key. Use a wallet such as MetaMask to generate a public-private key pair to ensure sufficient randomness. We have communicated with physicists who specialize in quantum computing, and in the future, it may be possible to access quantum hardware wallets to meet the security requirements for randomness. At least for now, using mainstream wallets like MetaMask to generate and manage private keys is relatively safe.
Never reveal your private key (including the mnemonic phrase) to anyone. Note the difference between private keys and public keys. The public key is your wallet address, which is your identity on blockchain, and you can safely share it with others. Private keys should never be shared with anyone.
Never forget your private key. Generally, a wallet can restore your private key based on your mnemonic phrase, so keep your private key or mnemonic phrase in a safe place. If you forget your private key and mnemonic phrase, you will not be able to recover it.

If you are concerned about security issues, you can avoid depositing too many assets in the blockchain wallet. Currently, it seems that depositing a few tens of dollars is sufficient for using DeSci-21cm to meet academic needs for a long time.

In addition, we recommend that authors keep a local copy of the submitted folder in its original form. This way, even if the original copy on IPFS is lost, you can recalculate the CID and prove that you published the paper at a certain point in time by matching it with the corresponding timestamp on the blockchain. If you don’t mind the extra economic cost, you can also upload a copy on Filecoin in its original form to ensure the same CID.

Origin of the name

The 21cm spectral line is a radio signal emitted by hydrogen atoms during the reionization phase after the formation of the universe. It marks the end of the cosmic dark ages and is considered the first ray of light in the universe. We hope that DeSci-21cm can become the first serious DApp application of blockchain.

Future Plans and Discussions

Operating Plan

The MVP version of our DeSci-21cm DApp will initially create one preprint repository and three DeSci journals, which are planned to be the comprehensive journal “Future,” the professional field journals “Industrial Data Science” and “Decentralization.” The authors of “Industrial Data Science” and “Decentralization” can serve as temporary editors in the early stage, or they can personally operate and grow the journals, and the first article of DeSci will be published in “Decentralization.” We hope that “Future” will grow into an influential comprehensive journal like “Nature” and “Science” in the future, so we will be very cautious in selecting editors. We believe that the quality of a journal essentially depends on the level and standards of the editors and reviewers during the peer review process, rather than the publisher’s brand. Of course, top publishers and journals have accumulated reputation in history, and our development and operation of DeSci-21cm is not to replace traditional publishers, but to provide academics with another choice.

In addition, two other journals are planned, one is “PKU Space Science Review,” which is the major of the two authors of this article when they were in school; the other is “Complex System,” which has a very active community in China, but there is no mainstream journal for this field yet. We will communicate and discuss with scholars in these two fields, hoping to find suitable editors and make some growth attempts.

In addition, because there are some inconveniences in using blockchain in China (such as buying virtual currency), in the early stage of our operation plan, we may focus more on overseas markets, and then decide how to operate in China based on actual operational experience.

When we discussed the feasibility of the project with several professors in space physics, plasma physics, and planetary remote sensing during the project planning phase in early 2022, they all mentioned that non-native English-speaking researchers have a certain language disadvantage in writing and reading, which has to some extent affected their growth rate in the early stages. Our plan at that time also included a translation authorization contract, which authorized or certified the multi-language versions of authors’ papers, and recorded the translator’s contribution and automatically executed economic incentives due to the authentication function of blockchain, and the authors did not have to worry about premature disclosure of the content of their papers. However, at the end of 2022, ChatGPT emerged, which seems to be able to complete paper translation with machines, so we have temporarily shelved this plan, but we will continue to observe whether such demand scenarios exist, and decide whether to do corresponding development and operation in the next step.

Technical Plan

As an MVP version, there are still many user experience optimizations that can be made for DeSci-21cm, such as automatically sending emails to corresponding personnel when the paper status changes. Currently, the version mainly involves direct interaction between the front-end and the blockchain, with only a small amount of back-end. In the future, we will consider using more back-end systems to handle complex data and processes while ensuring the decentralized nature, in order to provide a better user experience.

We hope to share value with the journal and relevant academic communities in the future. A decentralized journal has value contributed by all participants, and it is worth researching how to design a reasonable algorithm to calculate the shares of authors, editors, and reviewers based on their contributions recorded on the blockchain, and distribute the value of the journal (including both direct profits and long-term potential value from the reputation improvement brought by the quality of papers) to those who contribute the most. This is a data science problem worth studying.

Some researchers have pointed out [5] that the current paper evaluation system also needs to be reshaped. The current evaluation system focuses more on innovation and neglects reproducibility, which is an important reason for some academic misconduct. Therefore, it is hoped to introduce machine learning and other models to balance innovation and reproducibility and conduct a more comprehensive evaluation. The development of large language models seems to make it possible for machines to automatically grade and categorize the quality of papers. These are all potential technical directions worth exploring.

In addition, we have not fully utilized all the advantages of decentralized storage. For example, when organizing files with Merkel DAG on each node’s IPFS, only one copy of identical file blocks with the same hash value is kept, and two roughly similar file copies will take up far less storage than twice the storage space. Whether this feature can be used for automatic plagiarism detection when submitting papers seems to be a technical issue worth considering.

Moreover, we need to decide whether to maintain our own IPFS gateway or integrate with Filecoin, reduce user payments, or maintain a network for storing papers in university libraries and laboratories. This is a technical and management issue that needs to be planned according to the actual operation of the DeSci-21cm project.

In terms of paper formats, we hope that the platform can support more formats in the future, such as executable code and virtual environments, and formats like jupyter notebook, to make it easier to share data, papers, and code. This is also a possible direction for platform construction.

Possible Issues

One of the possible obstacles that DeSci-21cm may encounter in the early stages is that young scholars may have a strong motivation to submit papers to traditional journals in order to accumulate academic reputation and promotion capital. This is because submitting a paper to a journal with a certain history and reputation can be beneficial in this regard. However, since academic ethics prohibit submitting the same paper to multiple journals, there may be a stronger motivation to choose traditional journals when a paper is finalized.

There are two essential reasons why submitting the same paper to multiple journals violates academic ethics:

This behavior is considered to occupy more scarce reviewing resources.
The copyright of papers published in traditional journals belongs to the publishers, and submitting the same paper to multiple journals can lead to copyright disputes.

For the second point, the copyright of the DeSci platform originally belongs to the authors themselves. Therefore, it is worth discussing the research ethics issue of whether a mechanism can be established with traditional journals, where if the author authorizes it, papers published on DeSci can transfer the copyright to other journals.

Platform and project address

The domain name of DeSci-21cm is www.desci-21cm.com. If the project goes smoothly, we will continue to use this domain name.

The contract code is located on the project page on GitHub: GitHub — stereoF/Dapp_21cm.

If you want to contract us: FirmamentSESS@gmail.com

Acknowledgments

Thanks to Professor He Jiansen, Professor Fa Wenzhe, Professor Liu Jian, Li Chenfang, Huo Ran, Wu Fei, Ruan Wenzhi, Cai Lei, A Ercha, and Li Jinxing for their valuable feedback and suggestions during the project planning stage. Thanks to Zhang Wenzhuo for the discussion. Thanks to Li Shunran for patiently answering frontend problems encountered during the development process. Thanks to ChatGPT for their assistance in development, writing, and translation processes.