The following is a guest post from Trevor Owens, Digital Archivist with the Office of Strategic Initiatives.
In From Records to Data: Its Not Just About Collections Any More, Leslie Johnston explained how she is increasingly seeing a need for librarians, archivists and curators to shift from thinking about their collections as sets of individual records, items or objects to seeing them as a corpus of data. Leslie mentioned that our Viewshare project was a way for organizations to start exploring this way of thinking about their collections. I thought I would take this chance to walk through a few examples of the kinds of insights that come from playing around with collection data in Viewshare.
Creating an Interface Teaches You About Your Collection. The intent behind the free Viewshare service is to make it as easy as possible for librarians, archivists and curators to create and share interfaces to digital collections. In practice we have found that the process of creating an interface with Viewshare can be an invaluable experience for better understanding a collection as a whole. That collection level view becomes a powerful way to identify data that requires remediation.
What we learn when we create an interface. There are at least two kinds of things we learn when we build an interface to a collection. First, seeing everything at once lets you see emergent patterns in a collection. Second, that big picture view helps bring little inconsistencies in the data of individual items into view.
Emergent patterns in collections as data
The image below is from a collection of trade cards from Fulton Street in Brooklyn New York. It is a really neat collection to explore and putting the cards on a map makes for an interesting way to navigate through each of the digitized cards. But it does more than that. As you can see in the image of the map below, it becomes instantly clear that the cards are more densely distributed on the left side of the map. Without seeing these cards displayed on a map it would have been very hard to have a sense of how they were distributed, but once we plot them on a map it becomes very clear.
Big Picture Views Highlight Inconsistencies in Data
The image below is of a facet in a collection which shall remain nameless. In this case, the collection had a facet that allows you to change your view to only include items associated with a particular state. When you have hundreds of records it is difficult to track down minor spelling inconsistencies in collection data. Can you see the problem in the image below?
When you display a field in Viewshare you very quickly see that one time Tennessee was misspelled. A minor inconsistency which might have remained hidden on the 300th record in a catalog, now sticks out like a sore thumb. That little number 1 taunts the librarian in us all to go back and remediate the data at the source.
We browse through records, stopping to examine them individually. This one-at-a-time experience is invaluable for developing a careful context sensitive understanding of a record. With that said, when we treat our records as date we can start to see different kinds of relationships. Ultimately, I would suggest that pivoting between engaging with the individual records and exploring those records as information inside a data set can be invaluable to stewards of cultural heritage collections.
I hope the few examples I have provided from Viewshare prompt some ideas for how you might use it to play with and explore cultural heritage collections as data sets.
Oh, and feel free to request a free Viewshare account for your institution.