I don’t know what to call these little things I’m making with JavaScript. I am reluctant to call them computer programs since I have been told that what I am doing is not real programming because it isn’t sufficiently complicated or something. So let me keep calling them Gizmos.

Anyway, when I think of a Gizmo it usually comes in three phases: I think of a search issue it would be nice to have a tool for, I find the resources that could make the tool, and then I work through all the steps that need doing and turn them into puzzles –  first I need to get THIS data, and then I need to evaluate it THAT way, after that it needs to go to THE OTHER API, etc. So it usually works out that each Gizmo is a series of puzzles, and if I solve all the puzzles I get a Web tool at the end as my prize.

As I’ve learned things and discovered more tools, I’ve found that the puzzle method works well when setting challenges for myself. Yesterday’s challenge was “Put three APIs together in a useful way.” I failed the challenge because Wikipedia’s data was too unformatted, but I finished Backyard Scholarship anyway. It’s available at https://searchgizmos.com/backyard/ and requires a free Data.gov API key.

What was SUPPOSED to happen was that you enter a Wikipedia figure’s name, and Backyard would get the birthplace of the person from the Wikipedia API, translate the place to a zip with the Zippopotam.us API, and finally use the Data.gov API to find all higher education institutions within a 30-mile radius of that location. The domain names of the institutions are then aggregated into a Google site: search.

But Wikipedia’s birthplace data is really weird (that’s a separate rant) so I ditched that API and made it so users can enter the birthplace themselves. (I suspect a future challenge will be figuring out a way to make Wikidata’s horrible P19 attribute useful somehow.)

The idea behind Backyard Scholarship is that higher education institutions often focus especially on famous people who were born nearby. By creating a Google site: search limited to only those institutions within a radius of a birthplace, that focused attention and research pops right out.  Here’s what the results for the default search, Mark Twain, look like:

Much more focus on local people and events then you’ll get with a more general “Mark Twain” site:edu search.

That’s not to say that a Backyard Scholarship search is better than a regular Google site:edu search. It’s not. Instead I think they’re complementary, because they provide completely different results. For example, let’s take Robert J Conley, a Cherokee who’s my favorite author of westerns next to Elmer Kelton. Here’s a Google search for “Robert J. Conley “site:edu search:

Information-rich, useful results. Now let’s do the same search with Backyard Scholarship, adding in Conley’s birthplace of Cushing, Oklahoma:

Also relevant results, but much more local!

Of course, you’re not limited to searching only by birthplace. Louisa May Alcott was born in Germantown Pennsylvania but is more commonly associated with Concord Massachusetts. If you search for her name and that city/state, you get a lot of interesting results from Boston University:

There are a couple of little bugs –occasionally Backyard insists it doesn’t recognize a zip code – but other than that it’s fun to play with.

I spend a lot of time thinking about how to intersect whatever information a searcher has – a name, a place – with commonly -used and -understood data points relevant to that topic (birthplace, location, occupation, dates, etc) with the intention of guiding and informing a search in a transparent way. It’s an interesting road to go down and I’m looking forward to evolving more complex puzzles.