Content aggregation

Figshare aggregates content from all Figshare powered repositories and makes it all searchable through https://figshare.com/search. You can use the facets to narrow your results by where the records are coming from. But what if you want to aggregate content in your own Figshare repository? This article provides an overview on how to do this.

Figshare makes it very easy for users to upload files, add metadata, and publish high quality digital records. But a repository manager may want to add records from elsewhere to create a catalog. These records will not have files attached to them, but will have metadata and should point to where a user can find the file(s). Figshare makes this possible through its Batch Management Tool and the API. Any metadata harvested from external sources can be formatted and uploaded as catalog records in your Figshare repository.

First a quick side note one the gold standard way to aggregate content. Content aggregation is hard. Data and papers affiliated with your institution are likely spread across many different repositories: funder repositories, discipline specific data repositories, publisher repositories, and even other institutional repositories. Finding these records and keeping their information up to date is the first hurdle to getting an accurate view of your research ecosystem (and a massive one). Current Research Information Systems (CRIS) or more simply Research Information Systems (RIMs) (like Symplectic Elements) are designed to look for and harvest bibliometric information connected to your institution, typically through the author affiliation information available in large databases (like dimensions.ai) along with other data sources. Integrating the CRIS/RIM with a repository provides an opportunity for the most accurate representation of all that your researchers are producing. The repository can hold non traditional research outputs that otherwise would not have a home, and the CRIS/RIM can harvest the metadata and for those records to form a comprehensive researcher profiles and accurate reporting dashboards. But what if you have a Figshare repository but no CRIS/RIM? You still have options!

Figshare's bulk record tools can help you aggregate content from other sources in your repository. When creating these records, it is important that each record is represented by the same metadata as the original record, points users to the original record, and uses the same persistent identifier as the original record. The benefits of aggregating content in this way are

1. Easier reporting: Use the Figshare dashboards or API to report on more of your institution's research outputs than just what has uploaded directly by researchers.
2. Increased discoverability: Figshare public pages are marked up to be indexed by common search engines and other services, like Google Scholar and Google dataset search. If you add a record from another repository to your Figshare repository, you are helping end users find that record
3. Ensure long term stewardship: By storing this information you have a record of your institution's outputs even if something catastrophic happens to the 3rd party that stored an original. For example, if a small publisher suffers a data loss, you at least know what papers from your institution's researchers were lost. And if you were storing open access versions, even better.

To create aggregated content in your repository, here are the steps you'll need to follow:

1. Harvest the metadata from the original record location. This might be a batch download of search results, download via an API endpoint, or set up an OAI-PMH harvest. The metadata should be in a structured format.
2. Format the metadata to match the fields available in your repository. If using the Batch Management Tool, download a sample of your metadata as a template. If using the API, follow the JSON formatting.
3. Note if the records have DOIs or Handles and be sure to include those. Figshare records require a persistent identifier (PID). To upload existing DOIs or Handles you will need to ask Support to enable this functionality for your repository. If the records do not have PIDs you may need to decide if it's appropriate to provide a DOI or Handle.
4. Add the PID with a title in the Related materials field and it's good to select the option to have the record called out on the public page.
5. Be sure to add the repository user id for each author affiliated with your institution. The harvested records will show up in their profiles.
6. You can add each record as a metadata only record or a linked file record using the PID. The latter is a nice way to make it obvious where someone should go for the file(s).

‍

If you’d like to hear more about our turn-key repository functionality please contact us at info@figshare.com

‍