Skip to content
This repository has been archived by the owner on Sep 26, 2019. It is now read-only.

Design

Jump to bottom
David Riordan edited this page Jan 22, 2015 · 8 revisions

Design Specifications

High Level Overview

Scribe is designed to be a point-and-click configurable tool enabling scientists, librarians, and others to create text-based structured data projects from digitized documents.

Experience Walkthrough

Example 1

Ken the Curator has an idea for a crowdsourcing project. He wants "The Crowd" to transcribe the full text of Shakespeare's sonnets.

Ken the Curator looks at Scribe, realizes it's probably not for him.

Scribe isn't about getting to the full text of a thing, it's about collecting selected elements that are relevant in a structured form. This can be ultimately for the improvement of metadata, or to create new standalone data sets compiled from across a corpus of materials.

Features

  • Framework for image-based crowdsourcing of text extraction tasks
  • Workflow for designing transcription tasks
  • Multi-stage workflows so users don't have to do all work in a single pass
  • A library of combinable widgets for structuring transcription workflows
  • User accounts [Based on third party account authentication]

Doesn't provide

  • Image hosting
  • User authentication (to be provided by a third party such as Zooniverse Login or OmniAuth)
  • User Discussion Service (to be provided by a third party such as Zooniverse Talk or Discourse)

Components

Scribe's architecture is deeply inspired by Zooniverse's Panoptes / Panoptes Frontend, their platform for self-service citizen science projects.

Project Setup & Workflow Creation Tool

NOTE: This is the most TBD part of the application

A Scribe project starts with an idea: "I have a stack of images, and I'd like to get some data out of the text they contain, but there's too many of for me to go through on my own."

A scribe project starts as a stack of organized images, called subjects. Each subject is part of a grouping, which can be part of a collection.

Think of it in terms of Zooniverse's data model for Old Weather, which is about transcribing historical ships logs for climate data:

Classification Example
Collection WWI
Grouping Ship (e.g. Jamestown (1876))
Subject Pages of ship logs (ordered by date)

When a user comes to your Scribe project and starts working, they'll

Datastore & API

Once the workflow

Client Transcription Site

  • 100% Database-driven
  • Made up of reusable and style-able UI components for classification, structuring, and transcription
  • Components designed for "Composability", the outputs from one UI element easily feed the API or other UI components

Homepage

Projects get a homepage

Other Client Site Features

  • Login agnostic - use Zooniverse ID's, use Facebook ID's, use nothing at all. If it talks oAuth (or a dead-simple login like Zooniverse), you can use it here.
  • Discussion agnostic - Very simple to jump from a task to talking about that task and that subject, using something like Zooniverse Talk or Discourse.

Project Admin Tool

Clone this wiki locally