On 28 September 2018, MDPI launched Hypothesis annotation on its journal Sustainability, enabling inline public annotation integrated with MDPI profile accounts. A few weeks later, this functionality was rolled out across all MDPI journals. We reached out to Martyn Rittman, Publishing Services Manager at MDPI, to learn more about this initiative, and brought him together with Hypothesis Director of Publishing Heather Staines to have a conversation about how publishers running their own annotation services fit into the wider Hypothesis mission.
Please tell us a bit about MDPI and your role?
Martyn Rittman: MDPI is an open access publisher. We started in 1996, initially as a repository for physical chemical samples. The first journal, Molecules, published the synthesis of the chemicals and made the papers freely available online. Now we have over 200 open access journals covering all research areas. Our focus is on fast, open dissemination of high quality research. We also run conferences and other initiatives, such as Scilit, a database covering all published literature. I joined MDPI as an editor before moving onto the production side and now run several projects covering improvements to our journals, new publishing initiatives (including Preprints.org), and author services.
What interested MDPI in annotation, and why was Hypothesis software selected for this integration?
Martyn: MDPI exists to ensure that research is effectively communicated. This is primarily through journals, but we want to encourage other channels as well. Commenting on and annotating papers are a part of this. We looked initially at building our own tool, but soon realized that Hypothesis had done the work already and has some really nice features. Being open source was a big incentive, since we can adapt the software to fit our particular aims and requirements.
MDPI decided to build and run its own annotation server rather than using Hypothesis hosted services. Can you tell us your main reasons for doing so?
Martyn: Part of our philosophy is to operate services ourselves where we can rather than relying on third parties. We have developed a lot of software internally to handle our websites, the journal editorial process, and production. It gives us a large amount of flexibility to adapt and innovate where we need to. The same applied when we were looking at annotation. We have different ideas for how it could develop and integrate with other services in the future. Running our own Hypothesis server makes it much easier to adapt when necessary.
How does the ability for publishers to run their own instance fit into the wider Hypothesis mission?
Heather Staines: Hypothesis was founded as an open-source project to help enable a conversation across all online content. We recognize that different organizations have different needs when it comes to integrating and storing annotations. For some, the ease of using Hypothesis hosting fits their needs. For others, our entirely open-source codebase enables publishers or other ventures to host their own annotation instances. We’re excited to see this vision become reality as MDPI launches their own Hypothesis server and pleased that their experience was relatively straightforward.
For publishers who want to explore running their own instance, what factors should they consider, where should they look for more information?
Heather: Publishers who are considering setting up their own annotation servers should be aware that it takes time, resources, and technical capacity. Beyond these considerations, a key drawback is that users won’t be able to annotate your content as part of a continuous practice they use with other scholarly content across the web. We will be addressing this in the future with an annotation client that can be logged in to multiple services simultaneously, so that users can always annotate any content with their service of choice. We advise speaking with us ahead of time — as MDPI did — to learn about some of the challenges and to be sure the results will meet the needs of your organization, researchers, and readers.
To learn more about integrating collaborative annotation with your publications, contact us and subscribe to news from Hypothesis.
Some of the functionality deployed by MDPI is a bit different than standard Hypothesis. Can you tell us a little bit about what’s different and why you made these changes?
Martyn: Two principles that affected how we deployed Hypothesis were to offer a low barrier of entry to users and to keep annotations open and public. To achieve the first of these, we decided that it was better to use existing MDPI accounts for annotations rather than asking users to set up new accounts specifically for Hypothesis. The annotation feature is part of a larger plan to recognize the contributions of researchers in various contexts. An MDPI-based account makes it easier to reuse the data for new tools at a later time.
Another change we made was not to include private groups. We believe that comments are worth making publicly. In this way, all readers can benefit from feedback that could help their understanding of the paper.
Apart from these, we changed very little from the standard Hypothesis implementation. The basic structure and function worked very well for us and was straightforward to implement.
Users may be wondering about Hypothesis capabilities to enable third-party accounts to be connected to regular Hypothesis accounts. Can you tell us more about how this works now and what is in store on this front in the future?
Heather: First, I want to draw a distinction between the use of third-party accounts and the running of a separate instance of Hypothesis with local user accounts, as is the case with MDPI. Publishers who want to incorporate single sign-on through OAuth can do so with the hosted Hypothesis service. For example, that is how we implemented integration with eLife.
Work is currently underway to enable third-party account users — either hosted by Hypothesis or elsewhere as in the MDPI example — to have the full experience possible with regular Hypothesis accounts. This includes the ability to create private groups in a third-party namespace with group activity pages (coming in 2019).
We believe it is important to support organizations who wish to run their own instance as well as users who want to store annotations in the service of their choice.
What’s up next for MDPI?
Martyn: We have a few exciting projects brewing. I can’t give away too many details, but they are focused on supporting research communities and individual researchers. We also have a lot of work to do maintaining our journals and ensuring that we cope with increasing numbers of submissions. Open access and open research continue to be areas of growth and MDPI aims to be a major contributor to both.
Thanks to both Martyn and Heather for providing more information on the MDPI integration and future developments in the annotation space!
A pioneer in scholarly open access publishing, MDPI has supported academic communities since 1996. Based in Basel, Switzerland, MDPI has the mission to foster open scientific exchange in all forms, across all disciplines. Our 204 diverse, peer-reviewed, open access journals are supported by over 35,500 academic editors. We serve scholars from around the world to ensure the latest research is freely available and all content is distributed under a Creative Commons Attribution License (CC BY).