Tackling Complex Problems with Crowdsourcing Solutions
Crowdsourcing can be a potent tool to solve highly complex problems. Researchers from business and academia are finding fascinating new ways to apply this technology. One example involves Haoqi Zhang
, an assistant professor of electrical engineering and computer science at Northwestern University.
Zhang designs social computing systems that promote desired behaviors and outcomes. His research spans crowdsourcing, social computing, human-computer interaction, artificial intelligence and machine learning. Much of his work is focused on engaging crowds and communities in problem-solving and collective action in order to advance new design processes. Zhang describes his programs as “crowdware” that enables users to crowdsolve, while also giving them feedback to indirectly manage their output.
One of Zhang’s crowdsourcing tools is Mobi, which collects information from many anonymous sources on the Internet to help plan custom trip itineraries. Mobi reads natural language to take requests, such as, “I want to spend the weekend in Chicago,” along with specific details such as, “I want to dine along Lake Shore Drive, catch a Cubs game at Wrigley Field, and visit the SkyDeck at Sears Tower.”
The request is crowdsourced out to users, who get incentives to create an itinerary in a collaborative workspace app. Mobi also includes a sidebar tool called Brainstream, which suggests to-do items for users to know which information is most important.
Another of Zhang’s crowdsourcing inventions is Cobi
, which “community-sources” from academic communities to plan academic conferences. Committee members and presenters can give their inputs about who should lead a particular session, or who has scheduling conflicts.
Cobi groups papers around common themes so authors can mark which papers are most relevant to their own. Conference chairs use a Web-based visual interface to sort out the various tracks and align the best schedules. Cobi also gives recommendations to conference attendees on the best sessions for their interest areas.
Mobi and Cobi are just two examples of Zhang’s extensive research at Northwestern. In a paper titled “Crowdsourcing General Computation
,” he explains that a key challenge of human computation is the effective and efficient coordination of problem solving. This includes balancing the delicate interplay between machine algorithms and human abilities.
“While simple tasks may be easy to partition across individuals,” Zhang writes, “more complex tasks highlight challenges and opportunities for more sophisticated coordination and optimization, leveraging such core notions as problem decomposition, subproblem routing and solution, and the recomposition of solved subproblems into solutions.”
With crowdsourcing, we are now able to untangle those more complex subproblems and reassemble their answers into the best possible solutions. That’s the exciting new power of “crowdware.”