I’m happy to report that a post of ours on Reactive and Multiplatform Crowdsourcing has been featured on the renowned CrowdResearch blog (Follow the Crowd).
The editorial and advisory team of the blog is an exceptional one, including:
- Michael Bernstein, Stanford University
- Alex Quinn, University of Maryland
- Ed H. Chi, Google Research
- Rob Miller, MIT
and many others.
The post puts the focus on the need of controlling the crowd in response to the quantity and timing of completed tasks; the quality of responses and task results; and the profile, availability and reliability of performers.
- crowdsourcing, addressed with a modeling approach
- social networking, as alternative means for outsourcing tasks on the web
- multi-platform integration for allowing flexibility in task allocation
- expertise finding for matching tasks to performers
- reactive rules for declaring events and triggers that control the crowd.
The nice outcome is that by applying our methods you get significant advantage in terms of time, effort and quality. Some examples are reported in the figure below, showing up to 50% increase in precision of results and saving of more than 20% of executions, simply by changing the reactive rules for calculating majority.
Our post is somehow a summary of our three papers on crowd sourcing, as reported also in my previous post in this blog, namely:
- Our WWW 2013 paper on Reactive Crowdsourcing (also available on the WWW site)
- Our WWW 2012 paper on Crowdsourcing search, which focused on exploiting social networks and crowdsourcing platforms for improving search
- And our EDBT 2013 paper on finding the right crowd of experts on social networks for addressing a specific problem