HumanAL: Calibrating Human Matching Beyond a Single Task
It addresses the issue of unreliable human labels for researchers and practitioners in data annotation, though it appears incremental as it builds on existing calibration methods.
The paper tackles the problem of human annotation errors by building behavioral profiles for annotators and using machine learning to calibrate their input, showing improved labeling quality across three matching tasks, including cross-domain settings.
This work offers a novel view on the use of human input as labels, acknowledging that humans may err. We build a behavioral profile for human annotators which is used as a feature representation of the provided input. We show that by utilizing black-box machine learning, we can take into account human behavior and calibrate their input to improve the labeling quality. To support our claims and provide a proof-of-concept, we experiment with three different matching tasks, namely, schema matching, entity matching and text matching. Our empirical evaluation suggests that the method can improve the quality of gathered labels in multiple settings including cross-domain (across different matching tasks).