Crowdsourcing and Scalable Active Learning

How can active learning focus human effort when crowdsourced data acquisition scales to millions of tasks?

Research on using active learning to reduce the cost of large-scale crowdsourced data acquisition.

The problem

Crowdsourcing can become prohibitively expensive when every item receives the same amount of human attention.

The approach

Use active learning to identify which labels or tasks are most informative and allocate crowd effort selectively.

Research on using active learning to reduce the cost of large-scale crowdsourced data acquisition.

Main contributions

• Research framing and system design
• Methods, implementation, and empirical evaluation
• Open research artifacts and scholarly dissemination