This website uses cookies.

Cookies are small text files held on your computer. They allow us to create the best browsing experience for you on our site. By using this website or closing this message, you are agreeing to our Cookies policy.

inqicon

START A CONVERSATION

Share your requirements and we'll get back to you with how we can help.

Fill the form below or use your

Send via

account

account

Thank you for submitting your request.
We will get back to you shortly.

Crowdsourcing for Kaboom with MTurk

  • Kaboom
  • Kaboom

Project Overview

The success of KaBOOM largely relies on identifying communities that require playgrounds. But there isn’t any data available in the U.S that can provide this information. KaBOOM has managed to find 93,000 locations that has playgrounds. This data, however, is not accurate and has to be verified. Verification cannot be done without the help of human intelligence. It is not practically possible for KaBOOM employees to verify each and every location. Hiring people from various communities and using them to verify this information is also tedious. For KaBOOM to check the accuracy of the playground information, people from various communities must voluntarily verify it. Leveraging Amazon's online crowdsourcing tool, QBurst built a solution to help Kaboom tackle this seemingly unsurmountable feat with community participation.

Our Solution

QBurst provided a solution for KaBOOM’s problem. We used Amazon Mechanical Turk, an online crowdsourcing tool to obtain information on playgrounds from the MTurk workers. We developed a web interface from which KaBOOM can post task requests to the MTurk system and review the completed work. The tasks are to verify playgrounds in the U.S. using Google Maps.

QBurst developed two scripts using MTurk API. The first script periodically checks the KaBOOM web interface for new tasks and automatically post them as HITs in the MTurk system. KaBOOM representatives will create tasks in the web interface. The second script identifies reviewable HITs in the MTurk system and lists them in the KaBOOM interface. KaBOOM representatives may review and accept or reject the completed work. The scripts execute automatically and work in the background as daemons. The web interface integrates Google Maps and makes it easy for the representatives to create tasks. This was developed using Django web framework.

End Result

KaBOOM’s visions as a nonprofit organization is shaping out well with more volunteers contributing in the MTurk community. With more information at hand, parents and policy makers will find it easy to plan on building playgrounds for their children.

Client

KaBOOM is a nonprofit organization that helps communities build playgrounds. KaBOOM sees decline of outdoor play as an important problem which according to them, is closely related to many issues such as obesity, anxiety, depression, fragmented communities, stunted social developments and violence.

KaBOOM envisions a place to play within walking distance of every child in the U.S. It offers tons of training materials and know-hows for communities to build playgrounds. Not limiting to resources alone, KaBOOM also partners with many organizations to gather and provide necessary funds for communities to build playgrounds.

Technologies

  • Django
  • Python
  • Cron Jobs
  • MTurk API and Boto library
  • Google Maps