Hindawi launches open source peer review system

 

This week, Hindawi is releasing a new peer review system that will debut on Bioinorganic Chemistry and Applications. The new platform is open source, developed as part of Hindawi’s collaboration with the Collaborative Knowledge Foundation (Coko). This release is a first step towards a network of open publishing infrastructure that Hindawi, Coko, and collaborating organizations will develop and share with the research community.

The video below gives a brief introduction to the philosophy behind our new system.

What we’ve built

This first release is a submission and review system which will support a single journal following an Editor-in-Chief-led workflow.

  • An administrator role sets up the initial editorial board positions and can provide support at every stage of the process.
  • Authors create accounts and submit manuscripts through a simple submission form.
  • The designated EiC conducts an initial screening before assigning promising submissions to a handling editor.
  • The handling editor makes a second assessment of the manuscript before deciding whether to invite external reviewers.
  • After receiving a certain number of reviews and soliciting revisions from the authors, the handling editor makes a final recommendation to the EiC.
  • The EiC makes a final decision on whether or not a manuscript should be published.

The system contains the necessary roles and email notifications to power this system, as well as an API and data model that allows integrations with external systems, such as Hindawi’s quality assurance system. For the technically inclined, the code for our new platform is available in this Gitlab repository. The Coko community as a whole is producing a great deal of amazing work, which you can browse in Coko’s public projects page.

Our philosophy was to take a horizontal slice of the review workflow and create a viable end-to-end system with minimal complexity. The system today is limited. It supports just a single workflow. Parts of the platform that will need to be configurable to support additional journals are still hard-coded. And while we’re ready to help others get started with the system, this still requires a significant level of technical skill. We have a long list of improvements we know the system needs before it can be a viable alternative to the existing commercial review platforms.

We are giving something back to the community that we hope will be adapted and reused by other publishers.

We felt it was important to put the system in front of users right away, as we continue to make progress. This ability to be more agile and make rapid improvements to the platform is one of the benefits of the new system. The initial release is a small step forwards, but we are proud of this release for a number of reasons.

First, we are giving something back to the community that we hope will be adapted and reused by other publishers. This is a first step towards our goals for more open scholarly communications infrastructure.

Second, from a technical perspective, the system is far more modern than our legacy platforms. It allows us to deliver improvements continuously, testing with users at every stage.  

Finally, the system is the product of revolutionary levels of collaboration between the organizations working within the Coko community. We’ve shared not only development resources but also a vision for how the publishing experience can be improved for researchers. 

Simplifying and standardizing peer review

Hindawi was one of the early publishers to transition fully to Open Access, but we’ve been slower to adapt our peer review model. A constraining factor has been the challenge of evolving publishing software to support advances in open science without introducing unmanageable complexity into the editorial process.

Our author and reviewer pool is diverse. We publish in many areas of STEM and medicine, interacting with researchers from all over the world: 76,000 authors, 11,000 editors, and 46,000 reviewers from 178 countries in 2017. A scalable publishing platform that works across all locations and disciplines requires careful tradeoffs between simplicity and flexibility.

In pursuit of the perfect platform, it’s easy to take customization too far. Each configurable variable in a peer review platform results in an exponential increase in the workflows the system supports. Standardization keeps the experience intuitive for researchers, limits cost and staffing complexity for publishers, reduces the cost of switching platforms, and helps prevent errors from creeping into the review process.

Being part of a community encourages convention over configuration. Click To Tweet

When we began to look at modernizing our publishing platforms last summer, we quickly decided we wanted to continue developing our peer review platform in-house. We did not want to become dependent on an external organization for a critical piece of infrastructure. We wanted to retain control of our publication data and our development roadmap. But, we also saw the advantages in being on a platform with other users. Wider usage would encourage the development of third party tools and integrations with external systems. Being part of a community encourages convention over configuration.

In September 2017, we committed to joining the Coko community and in January of this year we began development of the system we are releasing this week. We chose open source because it offered the right balance between controlling development while being part of a larger community.

We initially chose Coko because we saw the right balance of subject-specific and technical expertise. Quite quickly, we saw other benefits from working with eLife, University of California Press, and Coko’s other partner organizations.

The Coko community is at the radically open end of open source software communities. We share more than just a codebase. We collaborate daily on Mattermost. We meet as a community every few months to synchronize our projects. Many of us conduct all planning, research, design and development entirely in the open.

By working as a community we can develop software that satisfies a larger audience. Click To Tweet

Coko has brought together a diverse group of research organizations who share a vision for open infrastructure. By working as a community, we get a much wider sampling of use cases and requirements, and we can develop software that satisfies a larger audience. We collaborate on design and user experience and we push for standards across all our products. Improvements are shared openly, accelerating the support for open science best practices.

Our four principles of openness

We’ve written before about how we believe the walled gardens emerging in scholarly communications are dangerous to the open science movement. We want researchers to retain a healthy choice of where and how to publish.

We’ve put forward four principles of openness that guide our technology choices: open source, open data, open integrations, and open contracts. In conjunction, these four principles reinforce that publishing options should be driven by researcher satisfaction, not tradition.

With open source and open data, journals can move freely between publishers, while retaining their submission system and editorial workflow. Papers migrate easily between systems, keeping switching costs down. With open integrations, external systems can harvest those papers and use metadata about the papers to provide value-added services. And through open contracts, we avoid unnecessary secrecy, long commitments, or termination fees that create unhappy marriages.

The Coko community brings all of the above. A non-profit steward to ensure the code base remains open. Working groups on data standards and manuscript portability. An open API to power integrations and an open invitation to contributors to join the project. And all developed on open licenses to minimize barriers to collaboration.

Join the community

The challenge in developing a new peer review system isn’t making a system that works for a single journal. The difficulty comes in making choices about where a system should be flexible and how to provide an intuitive experience for researchers across all disciplines. The challenging part of the process starts now. We will need to make choices about where to invest and what to make configurable.

We plan to continue developing and extending the platform ourselves. We are also excited to see how other organizations take products developed in the Coko community and adapt them to new applications.

A small taste of our upcoming plans include:

  • Support for a full commercial publishing program at scale.
  • Semantic extraction and HTML-first workflows.
  • Data portability and simplified migration between systems.
  • Integrations with external services.
  • An end-to-end open source stack.

We won’t stop with peer review. We’ll be working on quality assurance, production, and hosting as open source projects. The Coko community is going to be a big part of this work going forward, to ensure what we build together is widely reusable. If you have any questions or you’d like to join the effort, get in touch at technology@hindawi.com.

 

Andrew Smeall
Chief Digital Officer
Hindawi

 


The text and images in this blog post are by Hindawi and are distributed under the Creative Commons Attribution License (CC-BY).