Data for this exhibit came first from DBpedia by SPARQL query. That data served as input for a ruby scraper to collect the URIs of the images on Wikipedia. The attribute images were then hand edited in N3, concatenated with cwm, and passed through Babel to provide Exhibit json.