paper-with-me

홈 › Papers

Are We Modeling the Task or the Annotator? An Investigation of Annotator Bias in Natural Language Understanding Datasets

2019-08-21 · IJCNLP 2019 11 · Mor Geva, Yoav Goldberg, Jonathan Berant

Crowdsourcing has been the prevalent paradigm for creating natural language understanding datasets in recent years. A common crowdsourcing practice is to recruit a small number of high-quality workers, and have them massively generate examples. Having only a few workers generate the majority of examples raises concerns about data diversity, especially when workers freely generate sentences. In this paper, we perform a series of experiments showing these concerns are evident in three recent NLP datasets. We show that model performance improves when training with annotator identifiers as features, and that models are able to recognize the most productive annotators. Moreover, we show that often models do not generalize well to examples from annotators that did not contribute to the training set. Our findings suggest that annotator bias should be monitored during dataset creation, and that test set annotators should be disjoint from training set annotators.

📄 PDF Abstract BibTeX arXiv:1908.07898

Code (2)

mega002/annotator_bias tf
nathanlesage/so-classifier pytorch

Tasks

DiversityNatural Language Understanding

Similar Papers 제목 키워드 기반

Toward Annotator Group Bias in Crowdsourcing

2021-10-08 · ACL 2022 5 · Haochen Liu, Joseph Thekinen, Sinem Mollaoglu, Da Tang 외

Crowdsourcing has emerged as a popular approach for collecting annotated data to train supervised machine learning models. However, annotator bias can lead to defective annotations. Though there are a few works investiga…

Modeling Annotator Preference and Stochastic Annotation Error for Medical Image Segmentation

2021-11-26 · Zehui Liao, Shishuai Hu, Yutong Xie, Yong Xia

Manual annotation of medical images is highly subjective, leading to inevitable and huge annotation biases. Deep learning models may surpass human performance on a variety of tasks, but they may also mimic or amplify the…

Image SegmentationMedical Image SegmentationSegmentationSemantic Segmentation

End-to-End Annotator Bias Approximation on Crowdsourced Single-Label Sentiment Analysis

2021-11-03 · ICNLSP 2021 11 · Gerhard Johann Hagerer, David Szabo, Andreas Koch, Maria Luisa Ripoll Dominguez 외

Sentiment analysis is often a crowdsourcing task prone to subjective labels given by many annotators. It is not yet fully understood how the annotation bias of each annotator can be modeled correctly with state-of-the-ar…

MisconceptionsSentiment Analysis

Reducing annotator bias by belief elicitation

2024-10-21 · Terne Sasha Thorn Jakobsen, Andreas Bjerre-Nielsen, Robert Böhm

Crowdsourced annotations of data play a substantial role in the development of Artificial Intelligence (AI). It is broadly recognised that annotations of text data can contain annotator bias, where systematic disagreemen…

Transformer-based Annotation Bias-aware Medical Image Segmentation

2023-06-02 · Zehui Liao, Yutong Xie, Shishuai Hu, Yong Xia

Manual medical image segmentation is subjective and suffers from annotator-related bias, which can be mimicked or amplified by deep learning methods. Recently, researchers have suggested that such bias is the combination…

DecoderImage SegmentationMedical Image SegmentationSegmentation+1