Building a Bridge Between Journal Articles & Research Data�                   

Eleni Castro

Institute for Quantitative Social Science (IQSS) �Harvard University�@thedataorg

Alex Garnett

Public Knowledge Project Simon Fraser University�@axfelix

IDCC 2014 - Feb 26, 2014

��Motivation

Photo: Jean Liu

Quotes for “Why?”

"Any moves towards data sharing are dependent upon the cooperation of journals.”*�– Sergiu Ghergina and Dr. Alexia Katsanidou �

*from European Political Science 2013: Data Availability In Political Science Journals

​

"The most immediate of these obstacles is the lack of a consolidated infrastructure for the easy sharing of data" - JORD Project results via EDaWaX blog

�

A team was assembled…

2 year Sloan Foundation grant (2012-2014):�

Public Knowledge Project (PKP)

• Simon Fraser University
• Stanford University (John Willinsky)�

Dataverse Network Project

• Harvard University’s Institute for Quantitative Social Science (IQSS) (Gary King & Merce Crosas)
• Micah Altman – Director of Research at MIT

Project Proposal

Who?: Address the needs of journals in addition to researchers and data managers.

What?: Enable journals to seamlessly manage the submission, review, and publication of data associated with published articles.

How?: Build the needed technology and create awareness / incentives regarding the importance of data sharing and preservation.

Why?: Help increase the replicability and reusability of research data.

Integrating Open Source Systems

We plan to do this by integrating two well-established open-source systems:

​

• Open Journal Systems (OJS) �[Willinsky 2005]�
• Dataverse Network �[King 2007; Crosas 2011]

​

Dataverse Network

A repository for research data that takes care of long term preservation and good archival practices, while researchers can share, keep control of and get recognition for their data.

Note: Major New Version 4.0 Coming Soon (Spring 2014)

Dataverse provides…

• Option for backups and replication of data in different locations (LOCKSS) so data is never lost.
• Re-format for long term accessibility so data never become obsolete.
• Extract Variable, Header (FITS) Metadata from datasets.
• Universal Metadata standards (DDI, DC, DataCite).
• Interoperability with other systems through standard protocols (such as OAI-PMH, APIs).
• Generates a DOI for permanent linking to datasets.

The Dataverse takes care of the archival infrastructure (“plumbing”) for you!

Open Journal Systems (OJS)

​

​

​

​

Open source journal management and publishing system created by PKP to expand & improve access to research.

Open Journal Systems (OJS)

​

​

​

​

Journals using OJS to publish >=10 articles per year. Credit: Juan Alperin, PKP

OJS provides…

✓ Technical infrastructure for an entire editorial management � workflow (submission, peer review, indexing).

✓ ‘Plugin' architecture (a la WordPress, Drupal, etc) to easily � add new features without the need to change the entire� core code base.

✓ Universal Metadata standards (expandable Dublin Core).

✓ Interoperability with other systems through standard� protocols (such as OAI-PMH, APIs).

✓ Option for backups and replication of journals in different� locations (LOCKSS) so access is never lost.

OK, so what is the integration going to do?

OJS plugin for:

Data + supporting files + metadata sent via API (SWORD-based)

to Dataverse

Harvard Dataverse

OJS Journal

Which Workflow?

Slide acknowledgement: Merce Crosas

Progress to-date

1. Compiled a list of potential journals (~100) that we can work with.
2. Publishers reviewed our plugin workflow and mockups pre-development.
3. Completed and released first version of plugin+API.
4. Small sample of journals tested + provided feedback of integration via survey.
5. Some of our partners have begun production testing.
6. Working on improvements to UI and metadata customization for v2 release

Demo

http://lib-hf1.lib.sfu.ca/ojs-2.4.3

Survey Findings

• Overall positive: Most testers plan to implement plugin!
• OJS doesn’t have the most eye-catching UI. Dataverse link & data citation must be prominent.
• But users excited for an OJS 3 + Dataverse 4 overhaul!
• Default dcTerms metadata defined in the SWORD spec is too limited for all of our users; the ability to customize fields is important.
• Plugin should not treat �all supplemental files as �Dataverse-worthy �(e.g., authors’ notes).

​

​

Photo: Flickr Commons

​

​

​

Next Steps (Spring 2014)

1. Publish best practices for data sharing policies.
2. Develop & release updated versions of plugin + Data Deposit API.
3. Continue to release all code and documentation :)
4. Expand on Collaboration.

Photo: Flickr Commons

Advantages to Integration

• Reference implementation so others can reuse and extend (extension of SWORD).
• Streamlining authors’ article and data deposit process.
• Permanent 2-way linking of the published article with its archived data.
• Increase visibility/access, and encourage data citation, replication and re-use.�

​

Photo: Flickr Commons

Thank you!

Contact

Alex Garnett - garnett@sfu.ca�Eleni Castro - ecastro@fas.harvard.edu��Project Website�http://projects.iq.harvard.edu/ojs-dvn

​

References

• Crosas M. A Data Sharing Story. Journal of eScience Librarianship 1(3), 173-179. 2013.
• Crosas, M., The Dataverse Network™: An Open-Source Application for Sharing, Discovering and Preserving Data, D-lib Magazine 17(1/2). 2011.
• King, G. An Introduction to the Dataverse Network as an Infrastructure for Data Sharing. Sociological Methods and Research 32(2), 173–199. 2007.
• Willinsky, J. Open Journal Systems: An example of open source software for journal management and publishing. Library Hi-Tech 23 (4), 504-519. 2005.

�

Photo: Flickr Commons