Making molehills out of mountains

1
Rapid Data Entry Supporting high-throughput
digitisation workflows in EMu
Laurence Livermore1 Alex Fell2, Muhammad Nadat2,
Andrew Brown2 and Ben Sullivan2 1 The Natural
History Museum, London 2 KE Software, an Axiell
Group Company
2
The Digitisation Challenge

• Increased government and public expectation
• Aim to digitise 20 million specimens in 5 years
• Current CMS little provision for rapid data entry
• Need new tools to support digitisation

3
Solving the problem Rapid Data Entry (RDE)

• Browser-based interface for KE EMu
• Customisable apps
• Support rapid data entry
• Bulk record creation
• Field validation
• Normalise and atomise data
• Project-based approach

4
Project-based Digitisation

• Managed by one or more leads
• People may be members of more than one project
• Project information stored in the collections
  database
• Most projects will have multiple project-specific
  apps

5
Project Dashboard

• Permission dependent
• Three app categories
• Forms
• Editors
• Statistics
• Multiple apps support various stages/components
  of digitisation

6
Forms

• Creates new records, including label
  transcription
• Record sets can be filtered
• Filtered records are offered to
  editors/transcribers randomly
• Bulk editing and customised operations through
  scripts

7
Editors

• Global updater
• Resolve attachments
• Apply consistency
• More targeted than EMu global editor
• Also created by project lead

8
Statistics

• Simple reporting mechanism
• Based on record status
• Visualisation tool
• Bar chart
• Pie chart

9
Project Creation Administration

• Browser-based configuration
• Can reference any backend field
• Permissions can be set per users on both projects
  and apps

10
Example Project Workflow Botanical Sheets
Form 1 - Stub record creation from barcoded sheets
Form 2 - Transcription of localities and
collectors
11
Future RDE Development

• UX/UI improvements (desktop/tablets)
• Record navigation and management
• Ongoing improvements for NHMs digital
  collections programme
• Support for non-digitisation activities
• Statistics and reporting

12
Advantages Disadvantages

• Apps are very flexible
• No clientside installation required
• Display and customisation does not
  (necessarily) require core client modifications
• Steamlined field selection allows for rapid
  data entry
• Digitisation occurs directly into collections
  database means all data are in one place from
  creation through to
• Normalisation tools within collections database

Support for mobile/tablet devices allows
novel/unanticipated workflows /- Apps and record
sets need to be configured by a Data
Manager/super user /- Complex normalisation (of
complex data) requires desktop client - Requires
WiFi in collections areas - Mobile/tablets less
suitable for typing - Another system and
interface to support and maintain
13
New Workflows

• Applications outside of rapid digitisation
• Specimen relocation loans
• Condition checking collections audit
• Data capture from visiting scientists
• Crowdsourcing

Original photograph taken by John Cummings
http//bit.ly/1vTrzvk
14
Acknowledgements

• Management and testing
• Darrell Siebert, Annette Ure and testing staff
  (curators and data managers)
• Software development
• Alex Fell, Muhammad Nadat, Andrew Brown and Ben
  Sullivan (KE Software)