Speaker: Peter Organisiak (UIUC)
Title: Designing for Better Crowdsourcing: Improving the Quality of Crowd-based Data Collection
Information objects can be described in many ways. Having detailed metadata and a various people's interpretations of the object helps in providing better access and use. While collecting novel descriptions is challenging, crowdsourcing is presenting new opportunities to do so. Large-scale human contributions open the door to latent information, subjective judgments, and other encoding of data that is otherwise difficult to infer algorithmically. However, such contributions are also subject to variance from the inconsistencies of human interpretation.
This talk describes research on the problem of variance in crowdsourcing and investigates how it can be controlled both through post-collection modeling and better collection-time design decisions. The primary contribution of this work is an understanding of crowd data quality improvements from non-adversarial perspectives: that is, focusing on sources of error beyond poor contributors.
Peter Organisciak is a Postdoctoral Research at the HathiTrust Research Center. He holds a PhD in Library and Information Science from the University of Illinois, and a Masters degree in Humanities Computing. His work has been featured in New Scientist, won an Outstanding Contribution Award from the Canadian Society for Digital Humanities, and recieved paper awards from the Association for the Advancement of Artificial Intelligence (AAAI) and the Association for Information Science and Technology (ASIS&T).
Homepage: <a href="https://www.lis.illinois.edu/people/phd-students/peter-organisciak">http...
7:15am to 8:30am