Metadata, Tags, and Tropy

In previous posts, you may have noticed there are tags that categorize the post to make them searchable on my blog and to a lesser extent the greater world wide web. This isn’t by accident. In the digital humanities, much of the material that a digital humanist will work on is in digital archives, either official archives (like the Library of Congress) or digital archives created by the user. This week, we’re learning how to create our own digital archives using Tropy.

For this example, imagine you have a collection of hundreds of Magic the Gathering cards. You can just throw them all in a shoebox uncategorized and spends hours trying to find the exact card you were looking for next time you want to make a deck (or write an essay). Instead of doing this, you’d probably want to actually categorize your cards. You might get a box that has dividers, but then you’ve got to decide how to label the dividers: Do you want to label by color, mana cost, type of card, playstyle? Any type of categorizing could be appropriate depending on what type of deck you’re building. The categorization frustrates you, but eventually you decide to color code them and put the colors in order of manacost, it’s not perfect, but it will work for now. Next time you roll around to making a deck, you know what card you want to build it around, except, you remember the name, but not the color or mana-cost. The only thing you can do is go through every single card looking, because your archive wasn’t flexible enough for this situation.

I know this is a long metaphor, by stay with me here. After getting frustrated with your card archive, you dedicate an entire day to taking pictures of each of your cards and uploading them into Tropy, a free open-source digital archive desktop application. Now you get to categorize them in any way you want, and it will allow you to have the same card in multiple categories. So you can create metadata for each card:

Title: Calculating Lich
Creator: Antonio Jose Manzanedo
Date: 2019
Type: Creature – Zombie Wizard
Archive: Shoebox 2
Collection: Black Cards
Box: Creatures
Folder: 6 Mana
Identifyer: 003/064 M
Rights: Wizards of the Coast
Date Added: XXXX-XX-XX
Modified: XXXX-XX-XX

Then you can add the tags:  MTG, Lich, Black, Creature, Zombie, Wizard, Zombie Wizard, Menace, Flavortext, Rare, Antonio Jose Manzanedo, 2019, Wizards of the Coast, WOTC, 5Strength, 5Health, 6Cost, 4Colorless, 2Swamp, Aggro.

With this you can create sub-categories, for instance you can make a new “List” in your project, and call it “Strong Creatures,” something that doesn’t have to do with your tags directly, and you can add all cards you think should be included including the above calculated lich. You can have as many lists as you want. You can also click on any of the tags, like if you wanted to create a zombie deck, you could click on zombie and see every other card you’ve tagged as zombie.

The best part is: Tropy doesn’t interfere with any external organization you have. Tropy directly uses the image files on your computer, but doesn’t move or alter them: your files are untouched everywhere except in tropy. It also doesn’t inconvenience the physical collections either, the physical collections will still be organized however they are, but you’ve made it so much easier to both use the digital information, and eventually to find the card in the physical archives using the metadata of what box, folder, and identifier it is.

Now I know this was a long metaphor, or rather explanation with a strange example, but Tropy can be used for other projects, like taking pictures of documents and labeling them as letters, reports, essays, declarations, etc. from libraries and archives all over the world. Tropy is really versatile, and if you don’t like the metadata options available, there’s also support for adding your own custom metadata templates. You can write descriptions of the different files, include multiple pictures per object, transcribe documents in the software, and more. You can even get extensions to export your archives to other Roy Rosenzweig Center products like Omeka or Zotero.

For this week’s assingment I didn’t make an entire online archive of my Magic the Gathering card collection, but it will be an idea in my mind for the future. Instead I was asked to use the items I digitized for my digital kitchen project and give them metadata, tags, and descriptions. Here’s my small digital kitchen collection:

Now tags are really useful. I can already use it to sort a few categories, namely colors and “cookware” even with just a few items. Metadata is the trickier aspect, for a lot of these objects, the metadata is sparse. I don’t organize my kitchen like an archive, so I’m at a loss of what to put for box, folder, identifier, etc. Like I guess I could say that the pot is typically in the cabinet above the dishwasher on the right side of the bottom shelf, nested in other pots and pans of different sizes. That would be great, except there’s not an easy way to type that into such a small field. I could theoretically label different subsections of the kitchen into Alpha-numeric categories, so that cabinet could be cabinet B shelf 1, but there would need to be some sort of key in my kitchen or in the digital archive to explain which cabinet is which. Then there’s also the problem of these being regularly used tools, that pot isn’t on that shelf as I’m writing this, it’s currently in my dishwasher, so the metadata for box, folder, etc. wouldn’t help in this situation anyway.

There are some other aspects of metadata that are not quite right either. Since this collection is of pictures of items in my kitchen in order to create a digital kitchen, it’s missing a lot that would really lend to the digitization aspect. The first thing that comes to mind is that the meta-data includes the dimensions of the picture, but not the dimensions of the objects inside the picture. There’s a few things I could do to rectify this: if there’s only a few objects that need dimensions I could include it in the description, but the description isn’t as easily searchable as the metadata. I could also create a custom metadata template by going to Edit > Preferences > Templates > New Template. Even then, it’s not perfect, there aren’t boxes specifically for dimensions data, I would just have to include it in a normal string box.

Another metadata category that would be helpful for this collection is specifically the material. I’ve tried to include the material in the descriptions of each item, but the mug is ceramic, the chili pack is plasticized paper, the infuser is silicon, the soap is in a plastic spray bottle, and the pot is made of aluminum. I added the materials as tags for the items, but it would also be helpful to specifically have the material type for a collection made of different types of objects like this.

Nevertheless, Tropy is still a really helpful tool, and I can’t imagine not using it on future essays, (how did I not think of something like this when I was working on my essay about greek pottery last year?).

(Last Updated: 09/29/2022)