Home

Accurator is a crowdsourcing framework, allowing users to contribute their expert knowledge to cultural heritage collections. Accurator runs on top of the Cliopatria triple store. Having all data modelled according to the Linked Data principles and making the system localisable are key design choices in the development of the system.

System overview

The code in this repository is structured as a cpack with the following folders containing relevant code:

• api - contains a Prolog file with api definitions
• lib - Prolog files for backend functionality needed by the api
• rdf - RDF data files with schema definitions, interface labels and domain definitions
  • domain - folder containing files for various domains
• web - hosted files
  • css - Bootstrap css and Accurator css files
  • html - static html pages
  • img - logo, flags and background images
  • js - Bootstrap, JQuery and JavaScript files per page for front end logic

Dependencies

Accurator depends on other cpacks for the following functionality:

• annotation_service - api for saving annotations
• autocompletion - providing autocompletion functionality in the annotation templates
• cluster_search - search and recommendation algorithm
• media_cache - caches images from remote servers and resizes when necessary
• versioned_graph - stores the annotation made in a git repository

Wiki pages

In this Wiki we explain the following topics:

• Setup - how to setup your own Accurator instance
• Collection data - how to load collection data and select a subset for annotation
• Taxonomies - how to leverage a taxonomic structure in the annotation process
• Domains - how to setup your own annotation domain
• Interface adaption - define what to ask your user and adapt the interface to the domain
• Locale - translate the Accurator interface