How Does Beam Search improve Span-Level Confidence Estimation in Generative Sequence Labeling?

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Hashimoto, Kazuma, Naim, Iftekhar, Raman, Karthik
Format: Preprint
Veröffentlicht: 2022
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866909088615497728
author Hashimoto, Kazuma
Naim, Iftekhar
Raman, Karthik
author_facet Hashimoto, Kazuma
Naim, Iftekhar
Raman, Karthik
contents Sequence labeling is a core task in text understanding for IE/IR systems. Text generation models have increasingly become the go-to solution for such tasks (e.g., entity extraction and dialog slot filling). While most research has focused on the labeling accuracy, a key aspect -- of vital practical importance -- has slipped through the cracks: understanding model confidence. More specifically, we lack a principled understanding of how to reliably gauge the confidence of a model in its predictions for each labeled span. This paper aims to provide some empirical insights on estimating model confidence for generative sequence labeling. Most notably, we find that simply using the decoder's output probabilities \textbf{is not} the best in realizing well-calibrated confidence estimates. As verified over six public datasets of different tasks, we show that our proposed approach -- which leverages statistics from top-$k$ predictions by a beam search -- significantly reduces calibration errors of the predictions of a generative sequence labeling model.
format Preprint
id arxiv_https___arxiv_org_abs_2212_10767
institution arXiv
publishDate 2022
record_format arxiv
spellingShingle How Does Beam Search improve Span-Level Confidence Estimation in Generative Sequence Labeling?
Hashimoto, Kazuma
Naim, Iftekhar
Raman, Karthik
Computation and Language
Sequence labeling is a core task in text understanding for IE/IR systems. Text generation models have increasingly become the go-to solution for such tasks (e.g., entity extraction and dialog slot filling). While most research has focused on the labeling accuracy, a key aspect -- of vital practical importance -- has slipped through the cracks: understanding model confidence. More specifically, we lack a principled understanding of how to reliably gauge the confidence of a model in its predictions for each labeled span. This paper aims to provide some empirical insights on estimating model confidence for generative sequence labeling. Most notably, we find that simply using the decoder's output probabilities \textbf{is not} the best in realizing well-calibrated confidence estimates. As verified over six public datasets of different tasks, we show that our proposed approach -- which leverages statistics from top-$k$ predictions by a beam search -- significantly reduces calibration errors of the predictions of a generative sequence labeling model.
title How Does Beam Search improve Span-Level Confidence Estimation in Generative Sequence Labeling?
topic Computation and Language
url https://arxiv.org/abs/2212.10767