Predicting the Quality of New Contributors to the Facebook Crowdsourcing System

Neural Information Processing Systems: Crowdsourcing and Machine Learning Workshop


We are interested in improving the quality and coverage of a knowledge graph through crowdsourcing features built into a social networking service. In this setting, most participants are casual users, making only a few contributions, and do so incidentally in the course of using the service. Techniques that make assumptions about the matching of users to questions, or the number of answers per user or per question do not work well under such circumstances.

We present an approach to model user trust when prior history is lacking, so that we can incorporate more new users’ contributions into crowdsourced decisions, and provide quicker feedback to new participants. Specifically, we present a logistic regression classifier for first-time contributions, and study the effect of prior knowledge about user demographics on this classifier using Facebook crowdsourcing datasets.

Latest Publications

Sustainable AI: Environmental Implications, Challenges and Opportunities

Carole-Jean Wu, Ramya Raghavendra, Udit Gupta, Bilge Acun, Newsha Ardalani, Kiwan Maeng, Gloria Chang, Fiona Aga Behram, James Huang, Charles Bai, Michael Gschwind, Anurag Gupta, Myle Ott, Anastasia Melnikov, Salvatore Candido, David Brooks, Geeta Chauhan, Benjamin Lee, Hsien-Hsin S. Lee, Bugra Akyildiz, Max Balandat, Joe Spisak, Ravi Jain, Mike Rabbat, Kim Hazelwood

MLSys - 2022