Guardado en:
Detalles Bibliográficos
Autores principales: Scott, Michael, Liang, Siyu, Wassink, Alicia, Levow, Gina-Anne
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:https://arxiv.org/abs/2510.22495
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866918171433238528
author Scott, Michael
Liang, Siyu
Wassink, Alicia
Levow, Gina-Anne
author_facet Scott, Michael
Liang, Siyu
Wassink, Alicia
Levow, Gina-Anne
contents This paper presents a systematic evaluation of racial bias in four major commercial automatic speech recognition (ASR) systems using the Pacific Northwest English (PNWE) corpus. We analyze transcription accuracy across speakers from four ethnic backgrounds (African American, Caucasian American, ChicanX, and Yakama) and examine how sociophonetic variation contributes to differential system performance. We introduce a heuristically-determined Phonetic Error Rate (PER) metric that links recognition errors to specific linguistically motivated variables derived from sociophonetic annotation. Our analysis of eleven sociophonetic features reveals that vowel quality variation, particularly resistance to the low-back merger and pre-nasal merger patterns, is systematically associated with differential error rates across ethnic groups, with the most pronounced effects for African American speakers across all evaluated systems. These findings demonstrate that acoustic modeling of dialectal phonetic variation, rather than lexical or syntactic factors, remains a primary source of bias in commercial ASR systems. The study establishes the PNWE corpus as a valuable resource for bias evaluation in speech technologies and provides actionable guidance for improving ASR performance through targeted representation of sociophonetic diversity in training data.
format Preprint
id arxiv_https___arxiv_org_abs_2510_22495
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle A Sociophonetic Analysis of Racial Bias in Commercial ASR Systems Using the Pacific Northwest English Corpus
Scott, Michael
Liang, Siyu
Wassink, Alicia
Levow, Gina-Anne
Computation and Language
This paper presents a systematic evaluation of racial bias in four major commercial automatic speech recognition (ASR) systems using the Pacific Northwest English (PNWE) corpus. We analyze transcription accuracy across speakers from four ethnic backgrounds (African American, Caucasian American, ChicanX, and Yakama) and examine how sociophonetic variation contributes to differential system performance. We introduce a heuristically-determined Phonetic Error Rate (PER) metric that links recognition errors to specific linguistically motivated variables derived from sociophonetic annotation. Our analysis of eleven sociophonetic features reveals that vowel quality variation, particularly resistance to the low-back merger and pre-nasal merger patterns, is systematically associated with differential error rates across ethnic groups, with the most pronounced effects for African American speakers across all evaluated systems. These findings demonstrate that acoustic modeling of dialectal phonetic variation, rather than lexical or syntactic factors, remains a primary source of bias in commercial ASR systems. The study establishes the PNWE corpus as a valuable resource for bias evaluation in speech technologies and provides actionable guidance for improving ASR performance through targeted representation of sociophonetic diversity in training data.
title A Sociophonetic Analysis of Racial Bias in Commercial ASR Systems Using the Pacific Northwest English Corpus
topic Computation and Language
url https://arxiv.org/abs/2510.22495