Announcing the $10,000 museum collections bounty

It’s the year of Everything, Everywhere, All at Once. Which is also the promise of open data.

For the museum-collections bounty we’re hoovering up collections data from museums everywhere so that we can view them all at once. Picasso’s prints, Muhammed Ali’s gloves, a banana duct taped to a wall — if it’s in a museum, out in front or buried somewhere in the back, it goes in our Dolt database.

Whether your heart leads you towards American expressionism, Native American artifacts, or Daoist manuscripts, we can all agree that more data is better. And with $10k on the line, you should feel extra motivated to get that Monet.

The schema

“For every problem there is a solution which is simple, clean and wrong.” — Henry Louis Mencken.

A single table is simple, clean, and technically wrong. While each row corresponds to a unique object, the "makers" (artists, creators, etc.) can be numerous. We decided to go with the flat schema adopted by the Met, where multiple makers can go in a single cell, separated by pipes "|". The idea is to strike a balance between form and usability.

CREATE TABLE `objects` (
  `object_number` varchar(50) NOT NULL,
  `institution_name` varchar(500) NOT NULL,
  `institution_city` varchar(500),
  `institution_state` varchar(500),
  `institution_country` varchar(500),
  `institution_latitude` float,
  `institution_longitude` float,
  `department` varchar(1000),
  `category` varchar(300),
  `title` varchar(1000),
  `description` varchar(10000),
  `current_location` varchar(500),
  `dimensions` varchar(5000),
  `inscription` varchar(4000),
  `provenance` varchar(4000),
  `materials` varchar(10000),
  `technique` varchar(200),
  `from_location` varchar(4000),
  `culture` varchar(2000),
  `date_description` varchar(500),
  `year_start` int,
  `year_end` int,
  `maker_full_name` varchar(3000),
  `maker_first_name` varchar(500),
  `maker_last_name` varchar(500),
  `maker_birth_year` varchar(2000),
  `maker_death_year` varchar(2000),
  `maker_role` varchar(2000),
  `maker_gender` varchar(500),
  `acquired_year` smallint,
  `acquired_from` varchar(200),
  `accession_year` smallint,
  `accession_number` varchar(500),
  `credit_line` varchar(4000),
  `image_url` varchar(2000),
  `source_1` varchar(2048) NOT NULL,
  `source_2` varchar(2048),
  PRIMARY KEY (`institution_name`,`object_number`)
)

On your marks

The database is pre-seeded with the collections at The Met. And on our README page I've provided a list of museums you can start with. Ask us questions on Discord if you have any.

On your marks, get set…

(Van) Gogh.