(Commercial) Automatic Speech Recognition as a Tool in Sociolinguistic Research

Nina Markl

(Commercial) Automatic Speech Recognition as a Tool in Sociolinguistic Research

Nina Markl

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

As speech datasets used in sociolinguistic research increase in size, laborious and time-intensive manual orthographic transcription is a challenge, limiting the amount of (transcribed) data which can be analysed. In this paper, I discuss the use of (commercial) automatic speech recognition (ASR) as a tool in sociolinguistic research in the context of a case study: the Lothian Diary Project. I describe the kinds of errors produced by two commercial ASR systems for British English within the broader context of algorithmic bias in ASR, and suggest some best practices when working with ASR in sociolinguistic work.

Original language	English
Title of host publication	University of Pennsylvania Working Papers in Linguistics
Subtitle of host publication	Selected Papers from NWAV 49
Publisher	Penn Graduate Linguistics Society
Number of pages	12
Volume	28
Edition	2
Publication status	Published - 19 Sept 2022
Event	49th meeting of New Ways of Analyzing Variation - The University of Texas, Austin, United States Duration: 19 Oct 2021 → 24 Oct 2021 Conference number: 49 https://www.nwav49.org/

Conference

Conference	49th meeting of New Ways of Analyzing Variation
Abbreviated title	NWAV 49
Country/Territory	United States
City	Austin
Period	19/10/21 → 24/10/21
Internet address	https://www.nwav49.org/

Access to Document

Commercial Automatic_MARKL_DOA05082022_VORFinal published version, 197 KBLicence: All Rights Reserved

https://repository.upenn.edu/pwpl/vol28/iss2/11/Licence: All Rights Reserved

Cite this

@inproceedings{e9a363cb31344e8380db4222866841fa,

title = "(Commercial) Automatic Speech Recognition as a Tool in Sociolinguistic Research",

abstract = "As speech datasets used in sociolinguistic research increase in size, laborious and time-intensive manual orthographic transcription is a challenge, limiting the amount of (transcribed) data which can be analysed. In this paper, I discuss the use of (commercial) automatic speech recognition (ASR) as a tool in sociolinguistic research in the context of a case study: the Lothian Diary Project. I describe the kinds of errors produced by two commercial ASR systems for British English within the broader context of algorithmic bias in ASR, and suggest some best practices when working with ASR in sociolinguistic work.",

author = "Nina Markl",

year = "2022",

month = sep,

day = "19",

language = "English",

volume = "28",

booktitle = "University of Pennsylvania Working Papers in Linguistics",

publisher = "Penn Graduate Linguistics Society",

edition = "2",

note = "49th meeting of New Ways of Analyzing Variation, NWAV 49 ; Conference date: 19-10-2021 Through 24-10-2021",

url = "https://www.nwav49.org/",

}

Markl, N 2022, (Commercial) Automatic Speech Recognition as a Tool in Sociolinguistic Research. in University of Pennsylvania Working Papers in Linguistics: Selected Papers from NWAV 49. 2 edn, vol. 28, 11, Penn Graduate Linguistics Society, 49th meeting of New Ways of Analyzing Variation, Austin, Texas, United States, 19/10/21. <https://repository.upenn.edu/pwpl/vol28/iss2/11/>

TY - GEN

T1 - (Commercial) Automatic Speech Recognition as a Tool in Sociolinguistic Research

AU - Markl, Nina

N1 - Conference code: 49

PY - 2022/9/19

Y1 - 2022/9/19

N2 - As speech datasets used in sociolinguistic research increase in size, laborious and time-intensive manual orthographic transcription is a challenge, limiting the amount of (transcribed) data which can be analysed. In this paper, I discuss the use of (commercial) automatic speech recognition (ASR) as a tool in sociolinguistic research in the context of a case study: the Lothian Diary Project. I describe the kinds of errors produced by two commercial ASR systems for British English within the broader context of algorithmic bias in ASR, and suggest some best practices when working with ASR in sociolinguistic work.

AB - As speech datasets used in sociolinguistic research increase in size, laborious and time-intensive manual orthographic transcription is a challenge, limiting the amount of (transcribed) data which can be analysed. In this paper, I discuss the use of (commercial) automatic speech recognition (ASR) as a tool in sociolinguistic research in the context of a case study: the Lothian Diary Project. I describe the kinds of errors produced by two commercial ASR systems for British English within the broader context of algorithmic bias in ASR, and suggest some best practices when working with ASR in sociolinguistic work.

M3 - Conference contribution

VL - 28

BT - University of Pennsylvania Working Papers in Linguistics

PB - Penn Graduate Linguistics Society

T2 - 49th meeting of New Ways of Analyzing Variation

Y2 - 19 October 2021 through 24 October 2021

ER -

(Commercial) Automatic Speech Recognition as a Tool in Sociolinguistic Research

Abstract

Conference

Access to Document

Fingerprint

Cite this