Saved in:
Bibliographic Details
Main Authors: Han, Seong Woo, Adıgüzel, Ozan, Carpenter, Bob
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2405.19521
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866929553202479104
author Han, Seong Woo
Adıgüzel, Ozan
Carpenter, Bob
author_facet Han, Seong Woo
Adıgüzel, Ozan
Carpenter, Bob
contents In applied statistics and machine learning, the "gold standards" used for training are often biased and almost always noisy. Dawid and Skene's justifiably popular crowdsourcing model adjusts for rater (coder, annotator) sensitivity and specificity, but fails to capture distributional properties of rating data gathered for training, which in turn biases training. In this study, we introduce a general purpose measurement-error model with which we can infer consensus categories by adding item-level effects for difficulty, discriminativeness, and guessability. We further show how to constrain the bimodal posterior of these models to avoid (or if necessary, allow) adversarial raters. We validate our model's goodness of fit with posterior predictive checks, the Bayesian analogue of $χ^2$ tests. Dawid and Skene's model is rejected by goodness of fit tests, whereas our new model, which adjusts for item heterogeneity, is not rejected. We illustrate our new model with two well-studied data sets, binary rating data for caries in dental X-rays and implication in natural language.
format Preprint
id arxiv_https___arxiv_org_abs_2405_19521
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Crowdsourcing with Difficulty: A Bayesian Rating Model for Heterogeneous Items
Han, Seong Woo
Adıgüzel, Ozan
Carpenter, Bob
Machine Learning
In applied statistics and machine learning, the "gold standards" used for training are often biased and almost always noisy. Dawid and Skene's justifiably popular crowdsourcing model adjusts for rater (coder, annotator) sensitivity and specificity, but fails to capture distributional properties of rating data gathered for training, which in turn biases training. In this study, we introduce a general purpose measurement-error model with which we can infer consensus categories by adding item-level effects for difficulty, discriminativeness, and guessability. We further show how to constrain the bimodal posterior of these models to avoid (or if necessary, allow) adversarial raters. We validate our model's goodness of fit with posterior predictive checks, the Bayesian analogue of $χ^2$ tests. Dawid and Skene's model is rejected by goodness of fit tests, whereas our new model, which adjusts for item heterogeneity, is not rejected. We illustrate our new model with two well-studied data sets, binary rating data for caries in dental X-rays and implication in natural language.
title Crowdsourcing with Difficulty: A Bayesian Rating Model for Heterogeneous Items
topic Machine Learning
url https://arxiv.org/abs/2405.19521