Crowdsourcing and annotations

What's an Annotation?

In everyday use, an annotation is a note added to something. On the web, there is now a formal notion of an Annotation, provided by the Web Annotation Data Model.

Annotations link things together. Annotations are how we use the web to say things about other things on the web. The Web Annotation Data Model gives us a technical standard to link parts of objects to concepts, and other content on the web. They are how we extend the hyperlinked web into the archive.

When people participate in a crowdsourcing project, they create annotations. This might be tagging words in text or people in photographs, or writing transcriptions and comments, or putting markers on a map.

Enrichment produces annotations

Machines also create annotations. In our enrichment pipeline, text recognition and natural language processing create new information about digital objects by analysing the text and images they contain.

Our projects use the Web Annotation Data Model because that lets us standardise the way our components communicate about parts of resources. The same components - like Annotation Studio on the client or the Enrichment Pipeline on the server - can be reused in multiple projects.

Whether a project involves humans, like the National Library of Wales Crowdsourcing Project; or machines, like Chronicle 250, or both, like the Indigenous Digital Archive, we need a data store and tools for these annotations. This is an Annotation Server.

The Elucidate Annotation Server

Elucidate is our implementation of the Web Annotation Data Model and Web Annotation Protocol. In addition to the features defined by these standards, Elucidate has many extensions to support the kinds of query, access control, data load and export scenarios that emerge in real-world project work.

You can read more about the Elucidate API here.

Madoc

All of these elements come together in our Madoc platform, for building digital collections and running crowdsourcing projects on them.