Altmetric Blog

Best practices for tracking altmetrics for your digital library content

Stacy Konkiel, 10th October 2019

Digital libraries contain loads of important scholarly resources: digitized primary sources like letters and illuminated manuscripts, arts scholarship like images and videos, and even interactive, peer reviewed websites. Digital libraries (also known as “digital special collections”) are older than their institutional repository cousins, and also much more complex.

Unfortunately, many digital library websites are not optimized to share the basic metadata that allows Altmetric to find and follow conversations around their digitized objects like we can for journal articles and books. 

That means that the broader impacts of digital library content is difficult to understand at scale, which in turn makes it hard to make a case for funding, staffing, and other support for your digital library. 

Luckily, there are some general best practices that, once implemented, make it a snap for Altmetric to track your digital library’s content.

In this post, I’ll share some technical fixes for making it easier to find altmetrics for your digital library content, along with general resources on digital scholarship assessment—there are a lot of smart people working on other aspects of this challenge already!

What Altmetric needs to track your digital library: PIDs, metadata, and meta tags

In order for Altmetric to track content shared in a digital library, we need for you to have standard persistent identifiers like DOIs or Handles assigned to your digital library content, and for you to share that information along with basic metadata in your digital library’s webpage meta tags in a supported format

Altmetric uses this information to monitor our data sources for white-listed links to your digital library content, which we then clean and aggregate into reports on how your digital library content is being engaged with across the social web.

Institutions can then track the impact of their digital library’s content at scale as part of a larger Altmetric Explorer for Institutions subscription.

How altmetrics for digital library content works

Fig 1. An illustration of the imaginary Chihuahua Studies digital collection webpage for the techichi dog breed, based on the Outdoor Indiana digital collection produced by Indiana University, with content adapted from Wikipedia

Let’s look at an imaginary example: Sara, a digital humanist working in chihuahua studies, collaborates with Lilliput University to get a Chihuahua Studies website online (chihuahuastudies.lilliput.edu). The website displays digitized images of chihuahuas found in manuscripts, books, and newspapers over the centuries (e.g. chihuahuastudies.lilliput.edu/ancient/manuscript/techichi.html). 

Lilliput University is hip to what altmetrics services need to track altmetrics for the Chihuahua Studies website and the content it hosts. So, they’ve registered Crossref DOIs for the Chihuahua Studies webpage and each output it hosts, like this one…

URL Persistent Identifier
chihuahuastudies.lilliput.edu/ancient/manuscript/techichi.html http://doi.org/10.1234/76319

…and they’ve designed their website so output information is automatically shared on each webpage in a machine readable format. They put this information into the website meta tags, so that any person or machine that looks at the webpage source code for a page like “chihuahuastudies.lilliput.edu/ancient/manuscript/techichi.html” can easily gather essential information about the scholarly object being shared:

<meta name=”DC.creator” content=”Sara Smith” />
<meta name=”DC.identifier” content=”http://doi.org/10.1234/76319″ scheme=”DCTERMS.DOI” />
<meta name=”DC.title” content=”Ancient Good Boys: Techichi” xml:lang=”en” />

If Joe Burns, a dog breeder and chihuahua enthusiast, tweets a link to “chihuahuastudies.lilliput.edu/ancient/manuscript/techichi.html”, here’s how Altmetric identifies the mention and adds it to our database:

  1. Altmetric monitors Twitter for new links that include the domain name (“lilliput.edu”), which has been previously white-listed. 
  2. When we see a link shared from a whitelisted domain, our bots follow the full link and look at the webpage metatags to confirm that what’s being shared is a scholarly object (e.g. an image from an illuminated manuscript instead of an “About Us” webpage). 
  3. Once we’ve confirmed that the link points to a scholarly object, we then add Joe Burns’ Twitter mention to our database, and collate it alongside other mentions of the same scholarly object.
  4. The mention is then shared in the Details Page for the output.
Fig 2. An illustration of an Altmetric Details Page for a Chihuahua Studies item record.

This is the basic process by which altmetrics aggregators like Altmetric can track digital library content as easily as we track journals and books online. Any digital library that implements these technologies and contacts Altmetric to whitelist their domain can begin tracking mentions for their content within a matter of days.

That said, we appreciate that many digital libraries are hosted on platforms that do not make metadata available in webpage meta tags. Moreover, some digital libraries do not mint persistent identifiers for their content using supported standards like Handles or DOIs. While we’d like to be able to build custom solutions to track those digital libraries, we unfortunately cannot. That’s why we recommend that anyone looking to upgrade their digital library platforms consider using “out of the box” solutions like DSpace or Figshare, which share this data as a matter of course.

Digital library assessment: the bigger picture

Tracking engagement with your digital library’s content across the social web is just one part of a much larger digital library assessment toolkit. Other kinds of impacts might include cultural influence, collection reuse and repurposing, and use in educational contexts. For a general overview of the kinds of website data you might collect to understand your digital library’s use, check out the 2015 white paper, “Altmetrics and analytics for digital special collections and institutional repositories” or the blog post “15 types of data to collect when assessing your digital library”.

For a deeper dive, the single best resource for digital library assessment is the DLF Interest Group on Digital Library Assessment. In this group, experts from around the world share resources and recommendations for how to assess digital library impact, costs, metadata quality, and other topics. Check out the DLF Interest Group website and the IG’s interactive Google Group and DLF wiki for number of helpful assessment resources. 

Interested in using Altmetric Explorer for Institutions to track your institution’s digital library content? Email us at info@altmetric.com, we’d love to learn more!

Leave a Reply

Your email address will not be published. Required fields are marked *