Science of Collaboratories logo - link
 
 
 
An alliance to advance the understanding of collaboratories
Science of Collaboratories
   
   

Return to the list of Collaboratory Projects

 
   
 

Name of Collaboratory :

 

Clickworkers

 
 

Logo :

   
 
 

URL :

  http://clickworkers.arc.nasa.gov/top  
 

Collaboratory Status :

 
Completed   Start Date : 2000 End Date : 2001 Info Last Updated : Thu, Dec 1 2005 8:01pm PST
 
 

Primary Collaboratory Function :

  Open Community Contribution System  
 

Secondary Collaboratory Functions :

  Virtual Learning Community, Shared Instrument  
 

Domain(s) :

   
 

Brief Description of the Collaboratory :

 

Only an estimated 20% of the raw data collected in typical space missions is ever properly exploited. Mars Global Surveyor (MGS) arrived at the red planet five years ago and has sent home over 25,000 high-resolution images.
Dr. Bob Kanefsky at Nasa-Ames Research Center in California and a small group of researchers were inspired by the various downloadable "screensaver" programs that a computer can run to crunch data when it would otherwise be idle. The original of such screensavers is SETI@home. Also, astronomy has a long history of relyng on amateurs for data collection and interpretation. The Ames team developed Clickworkers, a piece of software downloadable from the NASA web site. Using the software the general public was invited to examine images from Mars made by the Viking Orbiter, and identify craters. "Clickworkers" downloaded Mars images obtained by the ongoing MGS project, circled each crater, measured its size, and classified it as being fresh, degraded, or a "ghost" (a heavily eroded or overlain crater, which must therefore be old).

With the rise of the Internet there is the possibility for a distributed network of human processing power. Between November 2000 and September 2001 more than 80,000 people marked nearly 2 million craters for measurement and classified the relative age of another 300,000. Their averaged results proved to be just as good as those produced by an expert crater raterówho would have had to dedicate several months of continuous effort to produce the same amount of data. Many participants spent a small amount of time on the site (thirty-seven percent of the work was done by one-time visitors) while others have spent many hours with the images. With the experience acquired by the latter group, their assessments improved as time went by. Also, by having several participants examine each image independently, the reliability and quality of the results were diagnosed.

This experiment on massive image analysis in plaentary science was possible because human beings' finely tuned perceptual abilities were far superior to the capabilities of the image-recognition algorithms of the time. Also, the distributed network of partcipants across the Internet were supported centrally via NASA Clickworkers web site, and required little guidance beyond what was offered on the web site.

 
 

Access to Instruments :

  Clickworkers downloaded images from the clickworkers web site and worked with the web interface to identify and classify craters on the surface of Mars.  
 

Access to Information Resources :

  Instructions were posted on the Clickworkers site on how to classify craters. Also, multimedia presentations on how craters age was also available.  
 

Access to People as Resources :

  An email address was provided for clickworkers. Fewer than 1% of the participants sent email but even this proved overwhelming for those answering it.  
 

Funding Agency or Sponsor :

 
National Aeronautics and Space Administration (NASA)
NASA Ames Research Center
 
 
 

Notes on Funding Agencies/Sponsors:
From Director's Discretionary Fund

 
 
 
TOTAL PARTICIPANTS:
 

Notes on Participants/Organizations:
85,000

   
     
 
 

Communications Technology Used :

  Internet, email  
 

Technical Capabilities :

  Asynchronous conversation
Email
 
  Key Articles :  

Benkler, Y. (2002). Coase's penguin, or, Linux and The Nature of the Firm Yale Law Journal, 112(3), 369-446.

Kanefsky, B., Barlow, N. G. & Gulick, V. C. (). Can distribuited volunteers accomplish massive data analysis tasks?. Retrieved December 31, 1969, from NASA Ames, University of Central Florida, NASA Ames Web site: http://clickworkers.arc.nasa.gov/documents/abstract.pdf

 
 

Project-reported performance data :

  The judgments of experts were compared to the Clickworkers, and statistical averaging of the latter were essentially as good as the former. Frivolous inputs were easily weeded out. Reports are available on the web site.  
 
         
    
  Home | About SOC | Workshops | Resources | News & Events  

University of California, Irvine Logo

University of California, Irvine

School of Information Logo

School of Information University of Michigan