Guest post by Jackie Proven – Increasing repository content at St Andrews using MERIT data.
January 10, 2011 1 Comment
In 2010 the University of St Andrews implemented a new Research Information System (RIS) called PURE and we began developing strategies for increasing the full text content in our repository. Like most institutions we use various techniques such as presenting at staff events and enabling administrators to make deposits. We knew that getting good quality research outputs from high profile staff would help champion the cause, and so our Library Director suggested looking at the database of RAE 2008 submissions produced by the MERIT project.
The fact that the metadata has been enhanced with publisher details and the main ROMEO conditions means it has proved to be a valuable resource. The database can be easily searched and downloaded for further analysis, and you can then target content in various ways. One possible workflow which I have tried is:
- Search the database and download as Excel file
- Deduplicate using ja_ID (journal article unique ID which is generated for each author)
- Add columns for author email and deposit date
- Filter by “PVersionPermittedInIr” (Publisher version you can deposit) to see articles that allow Publisher’s pdf
- Filter further by publisher and check ROMEO*
- Email relevant researchers with a bit of blurb about open access and offer to deposit these outputs on their behalf, suggesting at the same time they could add author versions of other publications.
- Deposit the full text (which in our case just means adding the file to the metadata already in PURE, and validate for transfer to the repository).
While it’s possible to find this data in other ways, the MERIT database is a nice tidy solution. Starting with RAE2008 data provides confidence when viewing your repository as a showcase. Working with a batch of outputs from one publisher certainly speeds up the copyright-checking part of the deposit process. (*Given ongoing discussion about the detail of publisher policies including the distinction between websites and repositories, some people may be happy to take the data at face value; some may want to read the ROMEO detail or full policy.)
We have also extended the workflow to look for other content using our RIS metadata. Having this visibility of authors, research outputs and related activity is very helpful to our advocacy efforts. As well as increasing content, it is proving to be a valuable way to start dialogue with our researchers.
For further details contact Jackie Proven, email: firstname.lastname@example.org