It’s day 3 of Dashboard Week at The Data School. Today we’re getting creative with The Metropolitan Museum of Art Collection API. Our challenge is to connect to this API and build a dashboard with the data we retrieved. The API contains datasets on over 470,000 artworks in the museum’s collection. With so many possibilities, I decided to retrieve data on artworks within the “Asian Art” department, eventually narrowing my scope to Japanese Scrolls.
Approach
I used Alteryx to connect to The Metropolitan Museum of Art Collection API, utilising the Objects endpoint. I wrote the following API call to retrieve a list of all artwork’s object IDs within the “Asian Art” Department:
API call in Postman app for Testing – https://collectionapi.metmuseum.org/public/collection/v1/objects?departmentIds=6
From here, it was just a matter of creating the API calls for each of the object IDs – this is where all the interesting information about the artworks comes in. However, this also where Alteryx almost broke down on me since there were many API calls to be made. I made the following workflow with an aim to efficiently extract the data, but it could definitely use some refining. I’d love to hear your feedback on how this could be made more efficient:
Insights
Since the topic of Japanese scrolls has a rich historic context, I started exploring the artists and eras from which the scrolls originated. I found that over 200 Japanese scrolls had an unknown, unidentified, or null artist. This was peculiar, so I compared this to a timeline of scrolls created in each century and found that most unknown scrolls come from the 14th century (the Muromachi Period). This contrasts the timeline of all scrolls where the majority come from the Edo Period. Upon further research, I found it was likely this reflects the socio-economic-political conditions in each era. Additionally, older scrolls are less likely to be attributed to artists possibly as a result of faded or difficult to read seals and damage from natural disasters and wartime.
Future Directions
Currently, due to time restraints and issues with the API call, my dashboard is in a semi-complete state. I want to continue exploring Japanese scrolls by looking into the scrolls themselves. The next section of my dashboard will aim to divide scrolls into meaningful categories, analysing The Mets collection to find trends that either reflect or clash with historical patterns.
Thanks for reading. Feel free to check out my dashboard over on Tableau Public. If you missed the last blog and want to keep up-to-date with Dashboard Week, check out my blog for Day 2 here.
