J-STAGE Data: evidence data platform for Japan’s learned society publishing
Keisuke Iida
Department for Information Infrastructure
Japan Science and Technology Agency
Japan Science and Technology Agency (hereafter JST) promotes research and development in Japan through funding basic research, commercialization of new technology, and promoting international collaboration. JST also provides a variety of information platforms and services, including J-STAGE, an electronic journal platform. In October 2019, JST commemorated the 20th anniversary of J-STAGE, which now hosts more than 3,000 journals, conference proceedings and other academic contents published in Japan. J-STAGE Data is a new data repository to make underlying data available for J-STAGE publications.
How did you come up with the idea of building J-STAGE Data?
Since its launch in 1999, we have invested significant resources in J-STAGE to keep it up to global e-journal standards and good practice, by adding new features from manuscript submission to peer review process to the dissemination of contents. Over these 20 years, however, the scholarly publishing environment has so rapidly evolved that we felt the need to revisit J-STAGE policies and operations in order to adapt to the changing standards and practice. We established our advisory committee in March 2018 to deliberate our mid- to long-term strategies. Their final report [available in Japanese only] boiled down strategic actions into three areas: updating the e-journal platform in response to new demands, strengthening the collaboration mechanism with Japanese learned society publishers, and optimizing the means for service quality improvements. Creating a data repository was part of the action plans reflecting recommendations in these three areas.
What were the changes that J-STAGE had to make?
The world of scholarly publishing has evolved in its technology with diversified contents over the years. The open access publishing model is widespread, and the emergence of preprints and other non-traditional research outputs no longer warrant a single platform just for peer-reviewed journals. Sharing underlying data for publications is a prevailing trend among scholarly publishers, underscored by increasing awareness of research ethics, transparency, and access needs for publicly funded research. In addition, long-term preservation of all the research products, as well as standardization of metadata and infrastructure, require us to constantly optimize our choice of technology and platforms.
How does J-STAGE Data meet those new demands?
J-STAGE already offered a service called “electronic supplement” to allow publishers to upload supplementary data for a journal article, but the number of files and their size offered was quite limited. Creating J-STAGE Data as a new, separate platform allowed us to leave technical legacies behind and incorporate new practices into our workflow. We adapted Figshare for Publishers as our base platform, which met our basic requirements such as DOIs for datasets, Creative Commons licenses, and a user interface to browse, search, and download. Since each dataset had to be associated with a corresponding J-STAGE publication, we also needed a review workflow in place as a data curation mechanism and also for the peer review process. Having a separate data platform made it more flexible to publish a wide range of file types in large sizes — including different contributors associated with each — and link everything back to the main J-STAGE publication. Basic usage data like views, downloads, and citations are important indicators to determine the success of J-STAGE Data; so far, we are seeing increased usage of these datasets compared to the supplementary data in the older format.
What were the challenges in creating J-STAGE Data and how did you overcome those challenges?
The initial challenge was how to accommodate Japan-specific requirements. Japan Link Center (JaLC) DOI was our default choice and the DOI minting process had to be customized in the Figshare platform. We added a number of metadata fields to allow both English and Japanese information for title, authors, descriptions, etc. Perhaps bigger challenges came outside of the system. We developed extensive user manuals for our J-STAGE users, who are quite familiar with the journal publishing process but not necessarily with data publishing. There are certain things you can do on the platform that are not in accordance with the J-STAGE publishing policies, so we had to come up with a standard data publishing workflow that does comply. This included the timing of data release, embargo setting, support communications, among other things. We developed our policies based on user feedback and after some trial and error.
Do you have any research data policies in Japan and how do those policies relate to J-STAGE Data?
Japan’s Cabinet Office has assembled the expert panel on open science since 2014 and published several recommendations. More broadly to cover the science and innovation policy, the national Basic Plan is renewed every five years and we are in the beginning of its 6th cycle (2021-2025). The latest plan calls for actions by government agencies and research organizations to refurbish research systems that form the backdrop of open science and data-driven research, where research data management and reuse is strongly encouraged.
JST established our open access policy in 2013, which has been replaced by the Policy on Open Access to Research Publications and Research Data Management that covers both OA publications and research data management. J-STAGE Data is an extension of an existing e-journal publishing platform and does not directly support our policy as a funding agency; however, since J-STAGE is widely used by Japanese society publishers, we believe J-STAGE Data can serve as a vital tool for those who are in need of a data publishing platform conforming to the national recommendations.
How was J-STAGE Data received by society publishers?
We spent the first year on a pilot basis with a small number of publishers on board. This soft launch was useful to gather feedback from early adopters and optimize our policies and operations. After J-STAGE Data was officially launched in March 2021, we started organizing hearing sessions for those publishers considering data publishing. We received mixed feedback partly due to the familiarity with data sharing practice in different fields and perhaps due to different levels of personal incentives, too. Some research disciplines may have a longer history of data sharing while other fields may have reservations due to the sensitivity of the datasets. As they come on board, the data curation process reveals different metadata needs by research fields and their practice. Overall, however, the response is positive and we are starting to receive more applications than we expected.
How do you plan to develop J-STAGE Data in the next few years?
Enriching the metadata is our next priority. For example, most datasets list “authors” of corresponding publications as data creators, which may not always be the case. We could more correctly capture each author’s (and others’) contributions if more granular metadata allow us to do so. Most datasets are labeled with a CC license and are openly published, but some datasets may require tighter access controls and an explicit copyright statement when necessary. Multi-language support is something we need to consider, too, as we expand our user base in Japan.
Finally, as we see more journals make use of J-STAGE Data for sharing evidence data for their publications, we hope to see a clear synergy between the two platforms. Usage increase is an obvious one, but it may also be possible for published data to inform new research more directly, in which case data citation will be a clear indicator and something we are keen to closely keep track of.