False Discovery Rate Control and Statistical Quality Assessment of Annotators in Crowdsourced Ranking
This provides a tool for improving data quality in crowdsourcing applications across fields like machine learning and sociology, though it is incremental as it adapts existing knockoff filters and algorithms.
The paper tackles the problem of annotator position bias degrading consensus labels in crowdsourced ranking by introducing a statistical framework to detect such bias and control the false discovery rate, with experiments on simulated and real-world data showing it ensures most discoveries are true and replicable.
With the rapid growth of crowdsourcing platforms it has become easy and relatively inexpensive to collect a dataset labeled by multiple annotators in a short time. However due to the lack of control over the quality of the annotators, some abnormal annotators may be affected by position bias which can potentially degrade the quality of the final consensus labels. In this paper we introduce a statistical framework to model and detect annotator's position bias in order to control the false discovery rate (FDR) without a prior knowledge on the amount of biased annotators - the expected fraction of false discoveries among all discoveries being not too high, in order to assure that most of the discoveries are indeed true and replicable. The key technical development relies on some new knockoff filters adapted to our problem and new algorithms based on the Inverse Scale Space dynamics whose discretization is potentially suitable for large scale crowdsourcing data analysis. Our studies are supported by experiments with both simulated examples and real-world data. The proposed framework provides us a useful tool for quantitatively studying annotator's abnormal behavior in crowdsourcing data arising from machine learning, sociology, computer vision, multimedia, etc.