Upgrade to Pro — share decks privately, control downloads, hide ads and more …

Human-Computer Collaboration at NYPL Labs

Human-Computer Collaboration at NYPL Labs

This talk is part of the Culture Analytics and User Experience Design Workshop organized by the UCLA's Institute for Pure & Applied Mathematics, April 11 - 15, 2016.

Abstract:
This talk will provide an insight into the Building Inspector, Oral History, and other projects developed at New York Public Library Labs, with an emphasis on design and HCI-related challenges and iteration processes. It will cover the rationale as to when NYPL Labs uses (or does not use) computational tools for data extraction and improvement, and how the library is learning to collaborate with its users on the creation of digital resources.

Mauricio Giraldo

April 14, 2016
Tweet

More Decks by Mauricio Giraldo

Other Decks in Design

Transcript

  1. human-computer collaboration
    mauricio giraldo arteaga
    @mgiraldo
    @nypl_labs
    IPAM Culture Analytics and User Experience Design, April 2016

    View full-size slide

  2. not a real library scientist

    View full-size slide

  3. flickr.com/photos/wallyg/6133216510

    View full-size slide

  4. access
    digitization metadata
    public
    traditional digital library program

    View full-size slide

  5. access
    digitization metadata
    public
    engagement
    r+d

    View full-size slide

  6. what happens after digitization?

    View full-size slide

  7. human-computer collaboration

    View full-size slide

  8. embrace imperfection
    corollary of “perfect is the enemy of good”

    View full-size slide

  9. « A designer’s definition of
    ‘perfect’ is different for
    computational designers. »
    because it is not achievable
    John Maeda

    View full-size slide

  10. human-computer collaboration

    View full-size slide

  11. computers are good at some things…

    View full-size slide

  12. Randall Munroe - xkcd.com/1140

    View full-size slide

  13. David Hagen - drhagen.com/blog/the-missing-11th-of-the-month

    View full-size slide

  14. people overestimate OCR quality

    View full-size slide

  15. okay… so maybe computers are not that good

    View full-size slide

  16. people are good at other things

    View full-size slide

  17. human-computer collaboration
    i avoid the term “crowdsourcing”

    View full-size slide

  18. two examples

    View full-size slide

  19. footprint
    material
    use type
    street names
    address
    floors
    name
    class
    geo location
    year
    skylights
    backyards

    View full-size slide

  20. like Google Maps for the 19th century
    but Google Maps cannot answer questions about the 19th century

    View full-size slide

  21. *this is a simulation. actual process is intensive. consult your mathematician before trying

    View full-size slide

  22. and now you start tracing those buildings by hand
    (˽°□°)˽Ɨ ˍʓʓˍ

    View full-size slide

  23. can we automate this?

    View full-size slide

  24. computers are good at some things…

    View full-size slide

  25. yay footprints!
    60,000+ of those!

    View full-size slide

  26. like OCR for maps!™
    (not really trademarked)

    View full-size slide

  27. but OCR is pretty bad
    ಠ_ಠ

    View full-size slide

  28. people are good at other things!

    View full-size slide

  29. people don’t choose to complete these

    View full-size slide

  30. we have over 60,000 footprints to check!
    will people want to do this?

    View full-size slide

  31. what is the minimum contribution we need?
    we want the lowest friction possible so people will want to contribute

    View full-size slide


  32. this was 2013, touch-screen mobile had taken off

    View full-size slide

  33. what about malicious users?
    or even well-meaning ones who make mistakes

    View full-size slide

  34. 75% or more agreement between 3 or more people
    arbitrary numbers that have worked for us

    View full-size slide

  35. YES is on the right side because most people are right-handed and the algorithm is right most of the time

    View full-size slide

  36. Building Inspector
    buildinginspector.nypl.org

    View full-size slide

  37. will people participate?
    remember that little tweet button?

    View full-size slide

  38. footprint
    material
    use type
    street names
    address
    floors
    name
    class
    geo location
    year
    skylights
    backyards

    View full-size slide

  39. check
    YES FIX
    address color fix
    *footprints marked as “NO” go to polygon heaven

    View full-size slide

  40. address
    had to use full keyboard on mobile because fractions

    View full-size slide

  41. we add new maps as old ones are completed
    the bottleneck now became geo-rectifying those maps ¯\_(ϑ)_/¯

    View full-size slide

  42. this is actually version 2

    View full-size slide

  43. (the magic of git)

    View full-size slide

  44. good tutorials are hard

    View full-size slide

  45. Super Mario Bros. (Nintendo, 1985)

    View full-size slide

  46. we have too many edge cases
    or: how i learned to stop worrying and embrace imperfection

    View full-size slide

  47. ¯\_(ϑ)_/¯
    people skip them anyway

    View full-size slide

  48. NYPL Community Oral History Project
    oralhistory.nypl.org

    View full-size slide

  49. make these stories more accessible

    View full-size slide

  50. mark
    transcribe

    View full-size slide

  51. by brian foo @beefoo

    View full-size slide

  52. allows for basic text search
    but it’s not a proper transcript

    View full-size slide

  53. we felt we needed something different

    View full-size slide

  54. computers are good at some things…

    View full-size slide

  55. like OCR for audio!™
    (not sure if they trademarked that)

    View full-size slide

  56. we get transcription “snippets”
    from 1 to about 6 seconds long in varying levels of quality

    View full-size slide

  57. people are good at other things…

    View full-size slide

  58. by brian foo @beefoo

    View full-size slide

  59. we conducted a few usability studies

    View full-size slide

  60. by brian foo @beefoo

    View full-size slide

  61. it’s hard to reach consensus
    ಠ_ಠ

    View full-size slide

  62. embrace imperfection

    View full-size slide

  63. transcribe.oralhistory.nypl.org

    View full-size slide

  64. transcribe.oralhistory.nypl.org

    View full-size slide

  65. made with customizability in mind

    View full-size slide

  66. storyscribe.themoth.org

    View full-size slide

  67. this is one week after launch

    View full-size slide

  68. it is still being improved

    View full-size slide

  69. two of several projects we’ve worked on so far

    View full-size slide

  70. of human-computer collaboration

    View full-size slide

  71. it’s a collaborative process
    Willa Armstrong, Shawn Averkamp, Paul Beaudoin, Brian Foo, Josh
    Hadro, Elizabeth Hummer, Ara Kim, Shana Kimball, Tom Listanti,
    Matthew Miller, Eric Shows, Bert Spaan, and more at NYPL…

    View full-size slide

  72. one more thing…

    View full-size slide

  73. lala.cursivebuildings.com

    View full-size slide

  74. how to decode the 3D data?
    in the browser

    View full-size slide

  75. stereo.nypl.org

    View full-size slide

  76. Boston Public Library Boston Public Library U.S. Geological Survey
    U.S. Geological Survey

    View full-size slide

  77. thank you!
    mauricio giraldo arteaga
    @mgiraldo
    @nypl_labs
    IPAM Culture Analytics and User Experience Design, April 2016

    View full-size slide