what i wantfromLinked DataDan Chudnov dchud at umich edu 2008-02-23code4lib 2009 preconf
part onewhat i care about
i care aboutcontentthat *lasts*
i care aboutpersistence of precisionresilience of recall
i care abouteliminatingsoftware
i care aboutdoing web stuffbetter
i don’t care aboutdata formats
i don’t care aboutthe LOD cloud
the “LOD cloud”a“ready reference”collection
the differenceisCollection Development
data fordata’s sakeselected datafor adesignatedcommunity
one isn’t usefulthe other is antiquated
makethisusefulturn thisintolinked data
part twothe obvious stepbut first a disclaimer
MY OPINIONSDO NOT REFLECTTHOSE OF:MY EMPLOYERTHE US GOV’TPRES. BARACK OBAMAMOST OTHER HOOSIERS
the obvious step
picture if you will
which, basically,was this
the webis made of links
but waitthere’s more
which, basically,was this
butnobody knows that
unless theyknow to look
so the obvious stepis
LINKTHISDATA
resource URIrel=contentsrel=contentsrel=contents(already has)resource URIresource URIresource URI
arrowsbecomecross-reference
it becomescrawl-ablemine-able
integrated accessthroughsmart crawl/index
precision improveswithrich metadataexplicitly linked
recall improveswithexplicit linksthrough authority recs
this isdoing web stuffbetter
i care aboutdoing web stuffbetter
just by doingHTML / HTTP
use URIs as names for thingsuse HTTP URIsprovide useful informationinclude links to other URIshttp://www.w3.org/DesignIssu...
needs nonew software
i care abouteliminatingsoftware
part threethe not obvious step
two problemswiththe obvious step
1. there’s more stuff
2. stuff goes away
lcsh.info’s demise:what did it break?
nothing
we’d only just startedthinkingwhat to do with it
what if we had?
what would havebroken?
links.links would havebroken
we’re not hereto learn about404’d data
if we applylinked data practicesto this
resource URIrel=contentsrel=contentsrel=contents(already has)resource URIresource URIresource URI
when an lcsh.infogoes awayall that breaks
same forallof thisbut notanyof this
a 404 webisn’t a better web
part fourwhat i thinkwe should work on
i care aboutcontentthat *lasts*
i care aboutpersistence of precisionresilience of recall
i wantlinked datathat lasts
*resilient*linked data
to-do #1cache linked data
if your site breakswhen links breakcache and link yourself
instead of this
do thisUChicagocachesame as
just likewe doin ILS/OPACs
but, yknow,the better web way
so if this happensyour links still work
to-do #2proxy linked data
if this happensyour cache still work
but you can’trestart
dvcs lesson:no central serveris better
make every cacheits ownlinked data source
instead of just thisUChicagocachesame as
enable thisUChicagocachesame assame assame assame assame as
if one goes downthe others live on
if others live onit’s all still linked
if it’s all still linkedwe can keep adding
so makeyour cached recsuseful linked URIs too
#1conventions forcaching linked data
#2conventions forproxying linked data
food for thought
but now,let’s eat!
what i want from linked data
what i want from linked data
what i want from linked data
what i want from linked data
what i want from linked data
what i want from linked data
what i want from linked data
what i want from linked data
what i want from linked data
what i want from linked data
what i want from linked data
what i want from linked data
what i want from linked data
what i want from linked data
what i want from linked data
what i want from linked data
what i want from linked data
what i want from linked data
what i want from linked data
what i want from linked data
what i want from linked data
what i want from linked data
what i want from linked data
what i want from linked data
what i want from linked data
what i want from linked data
what i want from linked data
what i want from linked data
Upcoming SlideShare
Loading in …5
×

what i want from linked data

439 views

Published on

108 slides on resilient linked data, specifically through proxying and caching. presented at the code4lib 2009 preconference on linked data in providence.

0 Comments
1 Like
Statistics
Notes
  • Be the first to comment

No Downloads
Views
Total views
439
On SlideShare
0
From Embeds
0
Number of Embeds
3
Actions
Shares
0
Downloads
8
Comments
0
Likes
1
Embeds 0
No embeds

No notes for slide

what i want from linked data

  1. 1. what i wantfromLinked DataDan Chudnov dchud at umich edu 2008-02-23code4lib 2009 preconf
  2. 2. part onewhat i care about
  3. 3. i care aboutcontentthat *lasts*
  4. 4. i care aboutpersistence of precisionresilience of recall
  5. 5. i care abouteliminatingsoftware
  6. 6. i care aboutdoing web stuffbetter
  7. 7. i don’t care aboutdata formats
  8. 8. i don’t care aboutthe LOD cloud
  9. 9. the “LOD cloud”a“ready reference”collection
  10. 10. the differenceisCollection Development
  11. 11. data fordata’s sakeselected datafor adesignatedcommunity
  12. 12. one isn’t usefulthe other is antiquated
  13. 13. makethisusefulturn thisintolinked data
  14. 14. part twothe obvious stepbut first a disclaimer
  15. 15. MY OPINIONSDO NOT REFLECTTHOSE OF:MY EMPLOYERTHE US GOV’TPRES. BARACK OBAMAMOST OTHER HOOSIERS
  16. 16. the obvious step
  17. 17. picture if you will
  18. 18. which, basically,was this
  19. 19. the webis made of links
  20. 20. but waitthere’s more
  21. 21. which, basically,was this
  22. 22. butnobody knows that
  23. 23. unless theyknow to look
  24. 24. so the obvious stepis
  25. 25. LINKTHISDATA
  26. 26. resource URIrel=contentsrel=contentsrel=contents(already has)resource URIresource URIresource URI
  27. 27. arrowsbecomecross-reference
  28. 28. it becomescrawl-ablemine-able
  29. 29. integrated accessthroughsmart crawl/index
  30. 30. precision improveswithrich metadataexplicitly linked
  31. 31. recall improveswithexplicit linksthrough authority recs
  32. 32. this isdoing web stuffbetter
  33. 33. i care aboutdoing web stuffbetter
  34. 34. just by doingHTML / HTTP
  35. 35. use URIs as names for thingsuse HTTP URIsprovide useful informationinclude links to other URIshttp://www.w3.org/DesignIssues/LinkedData.html
  36. 36. needs nonew software
  37. 37. i care abouteliminatingsoftware
  38. 38. part threethe not obvious step
  39. 39. two problemswiththe obvious step
  40. 40. 1. there’s more stuff
  41. 41. 2. stuff goes away
  42. 42. lcsh.info’s demise:what did it break?
  43. 43. nothing
  44. 44. we’d only just startedthinkingwhat to do with it
  45. 45. what if we had?
  46. 46. what would havebroken?
  47. 47. links.links would havebroken
  48. 48. we’re not hereto learn about404’d data
  49. 49. if we applylinked data practicesto this
  50. 50. resource URIrel=contentsrel=contentsrel=contents(already has)resource URIresource URIresource URI
  51. 51. when an lcsh.infogoes awayall that breaks
  52. 52. same forallof thisbut notanyof this
  53. 53. a 404 webisn’t a better web
  54. 54. part fourwhat i thinkwe should work on
  55. 55. i care aboutcontentthat *lasts*
  56. 56. i care aboutpersistence of precisionresilience of recall
  57. 57. i wantlinked datathat lasts
  58. 58. *resilient*linked data
  59. 59. to-do #1cache linked data
  60. 60. if your site breakswhen links breakcache and link yourself
  61. 61. instead of this
  62. 62. do thisUChicagocachesame as
  63. 63. just likewe doin ILS/OPACs
  64. 64. but, yknow,the better web way
  65. 65. so if this happensyour links still work
  66. 66. to-do #2proxy linked data
  67. 67. if this happensyour cache still work
  68. 68. but you can’trestart
  69. 69. dvcs lesson:no central serveris better
  70. 70. make every cacheits ownlinked data source
  71. 71. instead of just thisUChicagocachesame as
  72. 72. enable thisUChicagocachesame assame assame assame assame as
  73. 73. if one goes downthe others live on
  74. 74. if others live onit’s all still linked
  75. 75. if it’s all still linkedwe can keep adding
  76. 76. so makeyour cached recsuseful linked URIs too
  77. 77. #1conventions forcaching linked data
  78. 78. #2conventions forproxying linked data
  79. 79. food for thought
  80. 80. but now,let’s eat!

×