Today I got to do some investigating to answer the question…What do we do as an institution with disambiguating local names that go into our digital collections? If you work in digital initiatives, you must be nodding and thinking the same thing. How do you handle all of these different names that archivists, curators, work studies, other staff, and joe the friendly volunteer enter into the digital repository? I asked this question on the MODS listserv. I’ve also asked other colleagues about this. One of the most frequent answers is the spreadsheet or a text file. Both are low tech, don’t require any training necessarily, and anyone can access these files to see and check. These are great solutions. But I wanted to press the envelope a little. How do you have a solution to track names in the form you want to them to be in (hopefully always) in a non-proprietary format, accessible by everyone on your team or who needs this information, perhaps can have a unique ID associated with it, and where information in its records can be pulled and re-used for other projects (perhaps an autocomplete function). I’m just thinking quickly of out of the box features. Now the text file certainly gives us a non-proprietary format. However, not everyone can access this file. For example, if this text file lives on a desktop or a drive that can only be access through your desktop you might have problems accessing this file in certain circumstances. And of course, the names in the text file don’t have any id’s associated with it except those that perhaps you put in. This is the same problem with the spreadsheet.
Where I work, we can create name authority records in MARC in our ILS. Unfortunately, not everyone can access these records and it is mixed in with our other authority data which comes from a vended solution. It is also difficult to extract this information (we rely on one person, our systems librarian). To top things off, information is not readily searchable or even easy to discover with only a local id number assigned by the ILS. Essentially, name authority information for our digital initiatives would be buried and lost in our ILS.
That lead me to investigate another option. We use Fedora and Islandora for our digital repository. I created an Islandora collection (using just the small image content model or a collection that can only take jpg, gifs or pngs). The format or content type really didn’t matter since I wasn’t going to load a digital object but metadata about names. But since I was playing around, I decided to design a kitsch logo in jpg format to upload as well. That way, I didn’t have to see the ugly folder that is the default Islandora icon. Then I created a form where I could input essentially information for authority records and voila. Currently I only loaded 6 personal name authority records. They appear in alphabetical order. Thanks to our system, each record will be automatically assigned a handle; this is pretty cool since I can add this handle to MODS records in the valueURI attribute. It is accessible via the web anywhere, anytime (well except if the system is down).
I’m still playing with this whole idea. Essentially, my question was how to use my current resources to answer the questions of: non-proprietary format (the records are in xml), persistent or unique ID, accessible from wide variety of sources and places, be able to interact with our other information in our digital repository, and be able to reuse and share data. Granted, I haven’t solved all of these questions. Having data in Islandora is albeit open source but still tied to a number of conditions like any software. But it’s definitely a good start. Even if you don’t have Islandora, this is something you can do (namely create a collection with items that are only metadata records) in Omeka, DSpace or other digital library applications. What do you do with your names?