Fast, Not Fancy: Rethinking G2P with Rich Data and Rule-Based Models Article Swipe

PDF

Mahta Fetrat Qharabagh , Zahra Dehghanian , Hamid R. Rabiee ·

YOU? · · 2025 · Open Access · · DOI: https://doi.org/10.48550/arxiv.2505.12973

Homograph disambiguation remains a significant challenge in grapheme-to-phoneme (G2P) conversion, especially for low-resource languages. This challenge is twofold: (1) creating balanced and comprehensive homograph datasets is labor-intensive and costly, and (2) specific disambiguation strategies introduce additional latency, making them unsuitable for real-time applications such as screen readers and other accessibility tools. In this paper, we address both issues. First, we propose a semi-automated pipeline for constructing homograph-focused datasets, introduce the HomoRich dataset generated through this pipeline, and demonstrate its effectiveness by applying it to enhance a state-of-the-art deep learning-based G2P system for Persian. Second, we advocate for a paradigm shift - utilizing rich offline datasets to inform the development of fast, rule-based methods suitable for latency-sensitive accessibility applications like screen readers. To this end, we improve one of the most well-known rule-based G2P systems, eSpeak, into a fast homograph-aware version, HomoFast eSpeak. Our results show an approximate 30% improvement in homograph disambiguation accuracy for the deep learning-based and eSpeak systems.

Related Topics

Molly Jong-Fast

Fast X

The Fast And The Furious (2001 Film)

Fast Times At Ridgemont High

Angus Barbieri's Fast

Alexia Fast

Fast & Furious (2009 Film)

2 Fast 2 Furious

Fast Five

Fast & Furious 6

Thinking, Fast And Slow

Howard Fast

Concepts

No concepts available.

Metadata

Type: preprint
Language: en
Landing Page: http://arxiv.org/abs/2505.12973
PDF: https://arxiv.org/pdf/2505.12973
OA Status: green
OpenAlex ID: https://openalex.org/W4417286717

All OpenAlex metadata

Raw OpenAlex JSON

OpenAlex ID: https://openalex.org/W4417286717

Canonical identifier for this work in OpenAlex
DOI: https://doi.org/10.48550/arxiv.2505.12973

Digital Object Identifier
Title: Fast, Not Fancy: Rethinking G2P with Rich Data and Rule-Based Models

Work title
Type: preprint

OpenAlex work type
Language: en

Primary language
Publication year: 2025

Year of publication
Publication date: 2025-05-19

Full publication date if available
Authors: Mahta Fetrat Qharabagh, Zahra Dehghanian, Hamid R. Rabiee

List of authors in order
Landing page: https://arxiv.org/abs/2505.12973

Publisher landing page
PDF URL: https://arxiv.org/pdf/2505.12973

Direct link to full text PDF
Open access: Yes

Whether a free full text is available
OA status: green

Open access status per OpenAlex
OA URL: https://arxiv.org/pdf/2505.12973

Direct OA link when available
Cited by: 0

Total citation count in OpenAlex

Full payload

id	https://openalex.org/W4417286717
doi	https://doi.org/10.48550/arxiv.2505.12973
ids.doi	https://doi.org/10.48550/arxiv.2505.12973
ids.openalex	https://openalex.org/W4417286717
fwci
type	preprint
title	Fast, Not Fancy: Rethinking G2P with Rich Data and Rule-Based Models
biblio.issue
biblio.volume
biblio.last_page
biblio.first_page
is_xpac	False
apc_list
apc_paid
language	en
locations[0].id	pmh:oai:arXiv.org:2505.12973
locations[0].is_oa	True
locations[0].source.id	https://openalex.org/S4306400194
locations[0].source.issn
locations[0].source.type	repository
locations[0].source.is_oa	True
locations[0].source.issn_l
locations[0].source.is_core	False
locations[0].source.is_in_doaj	False
locations[0].source.display_name	arXiv (Cornell University)
locations[0].source.host_organization	https://openalex.org/I205783295
locations[0].source.host_organization_name	Cornell University
locations[0].source.host_organization_lineage	https://openalex.org/I205783295
locations[0].license
locations[0].pdf_url	https://arxiv.org/pdf/2505.12973
locations[0].version	submittedVersion
locations[0].raw_type	text
locations[0].license_id
locations[0].is_accepted	False
locations[0].is_published	False
locations[0].raw_source_name
locations[0].landing_page_url	http://arxiv.org/abs/2505.12973
locations[1].id	doi:10.48550/arxiv.2505.12973
locations[1].is_oa	True
locations[1].source.id	https://openalex.org/S4306400194
locations[1].source.issn
locations[1].source.type	repository
locations[1].source.is_oa	True
locations[1].source.issn_l
locations[1].source.is_core	False
locations[1].source.is_in_doaj	False
locations[1].source.display_name	arXiv (Cornell University)
locations[1].source.host_organization	https://openalex.org/I205783295
locations[1].source.host_organization_name	Cornell University
locations[1].source.host_organization_lineage	https://openalex.org/I205783295
locations[1].license
locations[1].pdf_url
locations[1].version
locations[1].raw_type	article
locations[1].license_id
locations[1].is_accepted	False
locations[1].is_published
locations[1].raw_source_name
locations[1].landing_page_url	https://doi.org/10.48550/arxiv.2505.12973
indexed_in	arxiv, datacite
authorships[0].author.id	https://openalex.org/A5114353445
authorships[0].author.orcid
authorships[0].author.display_name	Mahta Fetrat Qharabagh
authorships[0].author_position	first
authorships[0].raw_author_name	Qharabagh, Mahta Fetrat
authorships[0].is_corresponding	False
authorships[1].author.id	https://openalex.org/A5114353446
authorships[1].author.orcid
authorships[1].author.display_name	Zahra Dehghanian
authorships[1].author_position	middle
authorships[1].raw_author_name	Dehghanian, Zahra
authorships[1].is_corresponding	False
authorships[2].author.id	https://openalex.org/A5063512925
authorships[2].author.orcid	https://orcid.org/0000-0002-9835-4493
authorships[2].author.display_name	Hamid R. Rabiee
authorships[2].author_position	last
authorships[2].raw_author_name	Rabiee, Hamid R.
authorships[2].is_corresponding	False
has_content.pdf	True
has_content.grobid_xml	True
is_paratext	False
open_access.is_oa	True
open_access.oa_url	https://arxiv.org/pdf/2505.12973
open_access.oa_status	green
open_access.any_repository_has_fulltext	False
created_date	2025-10-10T00:00:00
display_name	Fast, Not Fancy: Rethinking G2P with Rich Data and Rule-Based Models
has_fulltext	True
is_retracted	False
updated_date	2025-12-12T22:30:26.305170
primary_topic
cited_by_count	0
locations_count	2
best_oa_location.id	pmh:oai:arXiv.org:2505.12973
best_oa_location.is_oa	True
best_oa_location.source.id	https://openalex.org/S4306400194
best_oa_location.source.issn
best_oa_location.source.type	repository
best_oa_location.source.is_oa	True
best_oa_location.source.issn_l
best_oa_location.source.is_core	False
best_oa_location.source.is_in_doaj	False
best_oa_location.source.display_name	arXiv (Cornell University)
best_oa_location.source.host_organization	https://openalex.org/I205783295
best_oa_location.source.host_organization_name	Cornell University
best_oa_location.source.host_organization_lineage	https://openalex.org/I205783295
best_oa_location.license
best_oa_location.pdf_url	https://arxiv.org/pdf/2505.12973
best_oa_location.version	submittedVersion
best_oa_location.raw_type	text
best_oa_location.license_id
best_oa_location.is_accepted	False
best_oa_location.is_published	False
best_oa_location.raw_source_name
best_oa_location.landing_page_url	http://arxiv.org/abs/2505.12973
primary_location.id	pmh:oai:arXiv.org:2505.12973
primary_location.is_oa	True
primary_location.source.id	https://openalex.org/S4306400194
primary_location.source.issn
primary_location.source.type	repository
primary_location.source.is_oa	True
primary_location.source.issn_l
primary_location.source.is_core	False
primary_location.source.is_in_doaj	False
primary_location.source.display_name	arXiv (Cornell University)
primary_location.source.host_organization	https://openalex.org/I205783295
primary_location.source.host_organization_name	Cornell University
primary_location.source.host_organization_lineage	https://openalex.org/I205783295
primary_location.license
primary_location.pdf_url	https://arxiv.org/pdf/2505.12973
primary_location.version	submittedVersion
primary_location.raw_type	text
primary_location.license_id
primary_location.is_accepted	False
primary_location.is_published	False
primary_location.raw_source_name
primary_location.landing_page_url	http://arxiv.org/abs/2505.12973
publication_date	2025-05-19
publication_year	2025
referenced_works_count	0
abstract_inverted_index.-	100
abstract_inverted_index.a	3, 61, 85, 97, 136
abstract_inverted_index.In	51
abstract_inverted_index.To	121
abstract_inverted_index.an	145
abstract_inverted_index.as	44
abstract_inverted_index.by	80
abstract_inverted_index.in	6, 149
abstract_inverted_index.is	16, 25
abstract_inverted_index.it	82
abstract_inverted_index.of	109, 127
abstract_inverted_index.to	83, 105
abstract_inverted_index.we	54, 59, 94, 124
abstract_inverted_index.(1)	18
abstract_inverted_index.(2)	30
abstract_inverted_index.30%	147
abstract_inverted_index.G2P	89, 132
abstract_inverted_index.Our	142
abstract_inverted_index.and	21, 27, 29, 47, 76, 157
abstract_inverted_index.for	11, 40, 64, 91, 96, 114, 153
abstract_inverted_index.its	78
abstract_inverted_index.one	126
abstract_inverted_index.the	69, 107, 128, 154
abstract_inverted_index.This	14
abstract_inverted_index.both	56
abstract_inverted_index.deep	87, 155
abstract_inverted_index.end,	123
abstract_inverted_index.fast	137
abstract_inverted_index.into	135
abstract_inverted_index.like	118
abstract_inverted_index.most	129
abstract_inverted_index.rich	102
abstract_inverted_index.show	144
abstract_inverted_index.such	43
abstract_inverted_index.them	38
abstract_inverted_index.this	52, 74, 122
abstract_inverted_index.(G2P)	8
abstract_inverted_index.fast,	110
abstract_inverted_index.other	48
abstract_inverted_index.shift	99
abstract_inverted_index.First,	58
abstract_inverted_index.eSpeak	158
abstract_inverted_index.inform	106
abstract_inverted_index.making	37
abstract_inverted_index.paper,	53
abstract_inverted_index.screen	45, 119
abstract_inverted_index.system	90
abstract_inverted_index.tools.	50
abstract_inverted_index.Second,	93
abstract_inverted_index.address	55
abstract_inverted_index.costly,	28
abstract_inverted_index.dataset	71
abstract_inverted_index.eSpeak,	134
abstract_inverted_index.eSpeak.	141
abstract_inverted_index.enhance	84
abstract_inverted_index.improve	125
abstract_inverted_index.issues.	57
abstract_inverted_index.methods	112
abstract_inverted_index.offline	103
abstract_inverted_index.propose	60
abstract_inverted_index.readers	46
abstract_inverted_index.remains	2
abstract_inverted_index.results	143
abstract_inverted_index.through	73
abstract_inverted_index.HomoFast	140
abstract_inverted_index.HomoRich	70
abstract_inverted_index.Persian.	92
abstract_inverted_index.accuracy	152
abstract_inverted_index.advocate	95
abstract_inverted_index.applying	81
abstract_inverted_index.balanced	20
abstract_inverted_index.creating	19
abstract_inverted_index.datasets	24, 104
abstract_inverted_index.latency,	36
abstract_inverted_index.paradigm	98
abstract_inverted_index.pipeline	63
abstract_inverted_index.readers.	120
abstract_inverted_index.specific	31
abstract_inverted_index.suitable	113
abstract_inverted_index.systems,	133
abstract_inverted_index.systems.	159
abstract_inverted_index.twofold:	17
abstract_inverted_index.version,	139
abstract_inverted_index.Homograph	0
abstract_inverted_index.challenge	5, 15
abstract_inverted_index.datasets,	67
abstract_inverted_index.generated	72
abstract_inverted_index.homograph	23, 150
abstract_inverted_index.introduce	34, 68
abstract_inverted_index.pipeline,	75
abstract_inverted_index.real-time	41
abstract_inverted_index.utilizing	101
abstract_inverted_index.additional	35
abstract_inverted_index.especially	10
abstract_inverted_index.languages.	13
abstract_inverted_index.rule-based	111, 131
abstract_inverted_index.strategies	33
abstract_inverted_index.unsuitable	39
abstract_inverted_index.well-known	130
abstract_inverted_index.approximate	146
abstract_inverted_index.conversion,	9
abstract_inverted_index.demonstrate	77
abstract_inverted_index.development	108
abstract_inverted_index.improvement	148
abstract_inverted_index.significant	4
abstract_inverted_index.applications	42, 117
abstract_inverted_index.constructing	65
abstract_inverted_index.low-resource	12
abstract_inverted_index.accessibility	49, 116
abstract_inverted_index.comprehensive	22
abstract_inverted_index.effectiveness	79
abstract_inverted_index.disambiguation	1, 32, 151
abstract_inverted_index.learning-based	88, 156
abstract_inverted_index.semi-automated	62
abstract_inverted_index.homograph-aware	138
abstract_inverted_index.labor-intensive	26
abstract_inverted_index.state-of-the-art	86
abstract_inverted_index.homograph-focused	66
abstract_inverted_index.latency-sensitive	115
abstract_inverted_index.grapheme-to-phoneme	7
cited_by_percentile_year
countries_distinct_count	0
institutions_distinct_count	3
citation_normalized_percentile