Crowdsourcing or bust: The Indexer, Archives NZ
- 1. Crowdsourcing or Bust Archives NZ’s ‘The Indexer’ Tracie Almond 23 April 2013
- 2. What is The Indexer? • A crowd-sourcing tool for ‘virtual volunteers’ to transcribe the contents of indexes from scanned images.
- 3. Archives NZ’s Project Objective • Improve the findability of archives that rely on secondary finding aids (e.g. name indexes). Supporting the long term vision that: • We have a single, integrated set of finding aids available online for public use.
- 4. Problems we’re trying to solve • Current search tools do not find some archives well due to the archives being described at a fairly lumpy level. • Names, places and other useful information are often only in indexes, etc. The indexes are often in paper form, so only available in our offices. • We are not serving our researchers as well as we could if we had that material incorporated into our online search tools. • We can then link to digital copies of the archive and reduce the need for people to visit our offices.
- 5. Archives NZ has about 1 million entries it would like to capture – like these ...
- 6. Class List – YCAF 4135 (example of Single entry type)
- 7. Hospital Index (example of Single entry type)
- 8. Auckland Hospital Charitable Aid – Application for Relief (example of Single & Multiple entry type)
- 9. BBCB 4243 Māori Succession Register (example of Multiple entry type)
- 10. The Indexer
- 14. Project benefits • High priority secondary finding aids will be accessible to our search engines. • Archives NZ will have the infrastructure needed to capture and make available many of the other high priority secondary finding aids. • We will increase our pool of volunteers and community involvement by providing online tools.
- 15. Lessons learned • Do your research – others have gone before and can provide great advice • Use a professional interface designer – it doesn’t cost much more to make it look nice and be a good user experience • Plan for openness – use authorities, make an API, open source the software (easier to do if you plan it in)
- 16. Do your research • Try out the other projects out there already – see what you do and don’t like • Talk to those who’ve done it – get their advice • Read up on the science behind it
- 17. Use a professional interface designer • It doesn’t cost much – less than $6k for The Indexer (including usability testing) • Test on users at the design stage and again before release • Result is tailored to the tasks but still clean
- 18. Plan for Openness • Can you use a metadata standard? • Can you adopt widely used name or place authorities? • Can you make an API to share the data? • Can you open source your software?
Editor's Notes
- #3: Note that project is not LIVE yet – date to be determined.
- #4: Data will feed search results for Archway.
- #9: More tricky with more data about the main applicant than the dependents.
- #10: Easy layout but tricky content – handwritten, Māori names
- #12: Home page: Intro/call to action Volunteers have some choices – topic and complexity Competition encouraged for those motivated by that Simple, clean design
- #13: Transcribing a simple index card
- #14: One of those tricky applications for relief
- #16: Budgeted <$6k for UI design / IA / usability testing Removes ‘friction’ and ‘anxiety’
- #18: Our designer is a star at the ‘information architecture’ as well as the graphic design side. Find one like that – not just someone who can make it pretty.
- #19: Failed pretty much on this whole thing other than can cross-walk the data to DC, via Archives NZ’s GAIMSII descriptive standard. Data is cross-walked to Archives NZ’s descriptive model, so it will map across to Archway.