Digital	
  Volunteers	
  &	
  
Crowdsourcing	
  to	
  Support	
  
Interna7onal	
  Relief	
  Efforts	
  
shadrock@ushahidi.com	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  @shadrocker	
  
Ushahidi	
  builds	
  tools	
  for	
  democra7zing	
  
informa1on,	
  increasing	
  transparency	
  and	
  
lowering	
  the	
  barriers	
  for	
  individuals	
  to	
  share	
  
their	
  stories.	
  
	
  
Open	
  Source	
  |	
  Non-­‐Profit	
  	
  
Community-­‐Driven	
  
 Common	
  Ques7ons	
  
•  What	
  /	
  who	
  are	
  “the	
  crowd”	
  or	
  “digital	
  
volunteers”?	
  
•  Organiza7onal	
  needs	
  
•  Accuracy	
  /	
  validity	
  of	
  data	
  
•  What	
  are	
  the	
  benefits	
  
 Example:	
  USAID	
  Data	
  Crea7on	
  
•  USAID’s	
  Development	
  Credit	
  Authority	
  
•  Targeted	
  Programming	
  
•  Analyzing	
  Impact	
  
•  BeXer	
  Coordina7on	
  
 Problem:	
  Data	
  were	
  broken	
  
COUNTRY LOCATION
Vietnam Mac Thi Buoi Vinh Tuy Ward, Hai Ba Trung Dist Ha Noi Viet Nam
Haiti Port au prince
Haiti Sud
COUNTRY LOCATION ADM1 PLACENAME
Vietnam Mac Thi Buoi Vinh Tuy
Ward, Hai Ba Trung Dist
Ha Noi Ha Noi
Haiti Port au prince Ouest Port au prince
Haiti Sud Sud Sud
	
  Solu7on:	
  data	
  cleanup	
  
!"	
  
	
  Step	
  1:	
  Infrastructure	
  Strategy	
  
!"	
  
!"!	
  !"	
  
!!	
  
"!	
  "	
  "
"	
  
!!
	
  Step	
  1:	
  Infrastructure	
  Strategy	
  
!"	
  
!"!	
  !"	
  
!!	
  
"!	
  "	
  "
"	
  
!!
	
  Step	
  1:	
  Infrastructure	
  Strategy	
  
•  Use	
  of	
  free	
  labor	
  
•  Non-­‐disclosure	
  
•  Personally	
  iden7fiable	
  informa7on	
  
•  Informa7on	
  Quality	
  Act	
  
	
  Step	
  2:	
  Pioneering	
  Policy	
  
 Step	
  3:	
  Volunteers	
  /	
  the	
  “Crowd”	
  
 Output:	
  Open	
  data	
  &	
  maps	
  
•  300	
  volunteers	
  
•  10,000	
  Records	
  
•  16	
  hours	
  
•  85-­‐90%	
  accuracy	
  
	
  Results:	
  Documented	
  Success	
  
•  Stronger	
  Public	
  Support	
  &	
  Interest	
  
•  Poli7cal	
  visibility	
  
•  Top	
  3%	
  most	
  viewed	
  USAID	
  web	
  pages:	
  
tools	
  &	
  data	
  can	
  obviate	
  organiza7onal	
  data	
  
silos	
  
	
  Impacts:	
  Documented	
  Success	
  
20%	
   15%	
  
•  Strategy	
  for	
  Technical	
  Infrastructure	
  
•  Pioneer	
  Policy	
  	
  
•  Build	
  “crowd”	
  with	
  volunteer	
  organiza7ons	
  
•  Clear	
  communica7on	
  with	
  volunteers	
  or	
  
“crowd”	
  is	
  cri7cal:	
  manage	
  feedback	
  loops	
  
•  Plan	
  on	
  accuracy	
  assessment	
  
•  Transparent	
  documenta7on	
  &	
  analysis	
  
	
  “Crowdsourcing”	
  Checklist	
  
•  OSM	
  is	
  a	
  open	
  map	
  (database)	
  of	
  the	
  world:	
  
users	
  create	
  data	
  by	
  tracing	
  satellite	
  imagery.	
  
•  Relief	
  organiza7ons	
  are	
  experimen7ng	
  	
  
•  Supported	
  by	
  U.S.	
  State	
  Dept.	
  “MapGive”	
  
•  Spa7al	
  accuracy	
  can	
  be	
  quite	
  good	
  (Haklay	
  et	
  
al.	
  find	
  >80%)	
  
•  Subjec7ve	
  Damage	
  Assessments	
  are	
  
problema7c	
  –	
  imagery	
  (American	
  Red	
  Cross)	
  
	
  Example	
  of	
  OpenStreetMap	
  
•  Kathmandu	
  Living	
  Labs	
  
•  Established	
  by	
  GFDRR:	
  Open	
  Data	
  for	
  
Resilience	
  
•  Built	
  local	
  capacity	
  to	
  create	
  open	
  data	
  
•  Mix	
  of	
  NGO	
  &	
  Crowdsourcing	
  
•  Valida7on	
  of	
  an	
  open	
  model	
  
	
  OpenStreetMap	
  in	
  Nepal	
  	
  
 OpenStreetMap	
  in	
  Nepal	
  	
  
	
  Kathmandu	
  Living	
  Labs	
  
We’re operating in an ecosystem now,
with no one that’s necessarily in charge.
The solution to journalism, civics, &
governance won’t be a single a-ha
moment, but the cumulative effect of
work done together in a network.”	
  
	
  
-­‐-­‐	
  Joi	
  Ito,	
  MIT	
  Media	
  Lab	
  Director	
  
“	
  

Know4 drr shadrock_roberts_may2015

  • 1.
    Digital  Volunteers  &   Crowdsourcing  to  Support   Interna7onal  Relief  Efforts   shadrock@ushahidi.com                      @shadrocker  
  • 2.
    Ushahidi  builds  tools  for  democra7zing   informa1on,  increasing  transparency  and   lowering  the  barriers  for  individuals  to  share   their  stories.     Open  Source  |  Non-­‐Profit     Community-­‐Driven  
  • 3.
     Common  Ques7ons   • What  /  who  are  “the  crowd”  or  “digital   volunteers”?   •  Organiza7onal  needs   •  Accuracy  /  validity  of  data   •  What  are  the  benefits  
  • 4.
     Example:  USAID  Data  Crea7on   •  USAID’s  Development  Credit  Authority   •  Targeted  Programming   •  Analyzing  Impact   •  BeXer  Coordina7on  
  • 5.
     Problem:  Data  were  broken   COUNTRY LOCATION Vietnam Mac Thi Buoi Vinh Tuy Ward, Hai Ba Trung Dist Ha Noi Viet Nam Haiti Port au prince Haiti Sud COUNTRY LOCATION ADM1 PLACENAME Vietnam Mac Thi Buoi Vinh Tuy Ward, Hai Ba Trung Dist Ha Noi Ha Noi Haiti Port au prince Ouest Port au prince Haiti Sud Sud Sud  Solu7on:  data  cleanup  
  • 6.
    !"    Step  1:  Infrastructure  Strategy  
  • 7.
    !"   !"!  !"   !!   "!  "  " "   !!  Step  1:  Infrastructure  Strategy  
  • 8.
    !"   !"!  !"   !!   "!  "  " "   !!  Step  1:  Infrastructure  Strategy  
  • 9.
    •  Use  of  free  labor   •  Non-­‐disclosure   •  Personally  iden7fiable  informa7on   •  Informa7on  Quality  Act    Step  2:  Pioneering  Policy  
  • 10.
     Step  3:  Volunteers  /  the  “Crowd”  
  • 12.
  • 13.
    •  300  volunteers   •  10,000  Records   •  16  hours   •  85-­‐90%  accuracy    Results:  Documented  Success  
  • 14.
    •  Stronger  Public  Support  &  Interest   •  Poli7cal  visibility   •  Top  3%  most  viewed  USAID  web  pages:   tools  &  data  can  obviate  organiza7onal  data   silos    Impacts:  Documented  Success   20%   15%  
  • 15.
    •  Strategy  for  Technical  Infrastructure   •  Pioneer  Policy     •  Build  “crowd”  with  volunteer  organiza7ons   •  Clear  communica7on  with  volunteers  or   “crowd”  is  cri7cal:  manage  feedback  loops   •  Plan  on  accuracy  assessment   •  Transparent  documenta7on  &  analysis    “Crowdsourcing”  Checklist  
  • 16.
    •  OSM  is  a  open  map  (database)  of  the  world:   users  create  data  by  tracing  satellite  imagery.   •  Relief  organiza7ons  are  experimen7ng     •  Supported  by  U.S.  State  Dept.  “MapGive”   •  Spa7al  accuracy  can  be  quite  good  (Haklay  et   al.  find  >80%)   •  Subjec7ve  Damage  Assessments  are   problema7c  –  imagery  (American  Red  Cross)    Example  of  OpenStreetMap  
  • 17.
    •  Kathmandu  Living  Labs   •  Established  by  GFDRR:  Open  Data  for   Resilience   •  Built  local  capacity  to  create  open  data   •  Mix  of  NGO  &  Crowdsourcing   •  Valida7on  of  an  open  model    OpenStreetMap  in  Nepal    
  • 18.
     OpenStreetMap  in  Nepal      Kathmandu  Living  Labs  
  • 23.
    We’re operating inan ecosystem now, with no one that’s necessarily in charge. The solution to journalism, civics, & governance won’t be a single a-ha moment, but the cumulative effect of work done together in a network.”     -­‐-­‐  Joi  Ito,  MIT  Media  Lab  Director   “