This is a guest blog post written by Seth Teicher summarizing the Human Computation Roadmap hosted by the Commons Lab here at the Wilson Center in June 2014. Seth is Head of Content and Business Development at CrowdFlower. He has worked in crowdsourcing his entire career — with Al Gore on a first-of-its-kind crowdsourcing initiative that leveraged tens of thousands of video testimonials to advocate for climate legislation, then as a founding team member at Atlas Obscura, a crowdsourced guide to hidden wonders of the world before joining CrowdFlower team in his current role.
“Human computation amplifies scientific discovery and human innovation.”
– Haym Hirsh, Cornell University
I had the honor of representing CrowdFlower at the Human Computation Roadmap Summit in Washington D.C. Led by Pietro Michelucci, author of the Handbook of Human Computation and founding editor of the journal Human Computation, and his co-chairs Lea Shanley of the Woodrow Wilson International Center for Scholars, Haym Hirsh, Dean of Computing and Information Sciences at Cornell University, and Janis Dickinson, Director of Citizen Science at the Cornell Lab of Ornithology, nearly 60 people from across academia, citizen science and the Federal government assembled to help sculpt a Human Computation Research Roadmap that aligns with our national priorities.
Putting Human Computation on the Political Map
From microtasking to game design, machine learning and beyond — we’d spend the next few days working in groups before presenting our ideas to Tom Kalil, Deputy Director of the White House’s Office of Science and Technology Policy. Tom would be a valuable proxy to prepare us for how best to frame the roadmap so it could gain traction and similar support to the newly launched National Robotics Initiative — an interagency program between the National Aeronautics and Space Administration, National Institutes of Health, National Science Foundation and the U.S. Department of Agriculture.
The Possibilities for Human Computation are Limitless
To orient ourselves to the depth and breadth of human computation and prepare for the brainstorming to come, we heard inspiring presentations from everyone in attendance. A few in particular were especially potent:
- Pablo Suarez, the associate director for research and innovation at the Red Cross / Red Crescent Climate Centre, led a participatory game to ground us in the harrowing decision points disaster planners face when integrating climate information into decision making.
- Professor Janis Dickinson of Cornell University’s Ornithology Lab discussed the widespread adoption of YardMap — a crowdsourcing initiative that empowers birdwatchers and homeowners to gather valuable data on avian habitats in their own backyard.
- Dr. Thomas Malone of MIT’s Sloan School and Center for Collective Intelligence shared some of the diverse ideas emerging from the Climate CoLab — an open innovation contest platform that solicits policy and product proposals that can move the needle on tackling climate change.
- We also learned about Stuart Lynn’s project, Zooniverse, in partnership with the Adler Planetarium, a suite of citizen science projects, most notably, Galaxy Zoo, that conscripts armchair astronomers to scour space telescope imagery to uncover our universe.
CrowdFlower and Social Impact
In the midst of these other presentations, I discussed CrowdFlower’s human-powered data enrichment platform that has been applied to a range of public interest data challenges:
- Text message categorization for disaster relief following the Haitian earthquake
- Unshackling election data in Indonesia
- Cellular photography analysis for breast cancer research at Harvard (See a sample of the CrowdFlower task used to annotate images for research below.)
How Can Microwork and Microtasking Shape Our World?
After delving into the stories of human computation, we were ready to get to work. The next day-and-a-half was spent in various working groups, composed of people from both technical and nontechnical backgrounds developing solutions ranging from games for climate adaptation (along the lines of this Tomogachi-like game submitted to the Climate CoLab) to ways in which online workers can leverage their microwork experience into real world credentials. (Get in touch with Markus Krause of Leibniz University to learn more about his microwork ideas.)
Securely Structuring Electronic Health Records
My group focused on securely structuring electronic health records (EHR) with CrowdFlower’s platform. Our idea coalesced around the transcription and structuring of the disparate EHRs currently languishing in our slowly-but-surely evolving healthcare system. It became clear that CrowdFlower could be an ideal tool to take on an important portion of this trillion dollar challenge in a scalable, secure, and cost-effective way. While complex, a workflow that feeds PDF records through a series of CrowdFlower jobs can unify these record both semantically and syntactically, facilitating greater efficiency of movement for EHRs between existing systems and emerging APIs. Effectively saving tremendous time and money for an issue of substantial national import.
Over the coming weeks, I look forward to collaborating further with my fellow attendees to advance the research roadmap and continuing to explore how CrowdFlower can support citizen science and public sector challenges.
How would you use CrowdFlower for social impact?
If you have ideas or initiatives that should be brought to our attention, or ways CrowdFlower could be leveraged for social impact, please drop me a line at email@example.com or on Twitter at @sethteicher.