ARTICLE: Annotator Reliability Through In-Context Learning

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Dutta, Sujan, Pandita, Deepak, Weerasooriya, Tharindu Cyril, Zampieri, Marcos, Homan, Christopher M., KhudaBukhsh, Ashiqur R.
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866913509163401216
author Dutta, Sujan
Pandita, Deepak
Weerasooriya, Tharindu Cyril
Zampieri, Marcos
Homan, Christopher M.
KhudaBukhsh, Ashiqur R.
author_facet Dutta, Sujan
Pandita, Deepak
Weerasooriya, Tharindu Cyril
Zampieri, Marcos
Homan, Christopher M.
KhudaBukhsh, Ashiqur R.
contents Ensuring annotator quality in training and evaluation data is a key piece of machine learning in NLP. Tasks such as sentiment analysis and offensive speech detection are intrinsically subjective, creating a challenging scenario for traditional quality assessment approaches because it is hard to distinguish disagreement due to poor work from that due to differences of opinions between sincere annotators. With the goal of increasing diverse perspectives in annotation while ensuring consistency, we propose \texttt{ARTICLE}, an in-context learning (ICL) framework to estimate annotation quality through self-consistency. We evaluate this framework on two offensive speech datasets using multiple LLMs and compare its performance with traditional methods. Our findings indicate that \texttt{ARTICLE} can be used as a robust method for identifying reliable annotators, hence improving data quality.
format Preprint
id arxiv_https___arxiv_org_abs_2409_12218
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle ARTICLE: Annotator Reliability Through In-Context Learning
Dutta, Sujan
Pandita, Deepak
Weerasooriya, Tharindu Cyril
Zampieri, Marcos
Homan, Christopher M.
KhudaBukhsh, Ashiqur R.
Computation and Language
Machine Learning
Ensuring annotator quality in training and evaluation data is a key piece of machine learning in NLP. Tasks such as sentiment analysis and offensive speech detection are intrinsically subjective, creating a challenging scenario for traditional quality assessment approaches because it is hard to distinguish disagreement due to poor work from that due to differences of opinions between sincere annotators. With the goal of increasing diverse perspectives in annotation while ensuring consistency, we propose \texttt{ARTICLE}, an in-context learning (ICL) framework to estimate annotation quality through self-consistency. We evaluate this framework on two offensive speech datasets using multiple LLMs and compare its performance with traditional methods. Our findings indicate that \texttt{ARTICLE} can be used as a robust method for identifying reliable annotators, hence improving data quality.
title ARTICLE: Annotator Reliability Through In-Context Learning
topic Computation and Language
Machine Learning
url https://arxiv.org/abs/2409.12218