Amharic Speech Search Using Text Word Query Based on Automatic Sentence-like Segmentation

<p dir="ltr">More than 7000 languages are spoken in the world today. Amharic is one of the languages spoken in the East African country Ethiopia. A lot of speech data is being made every day in different languages as machines are getting better at processing and have improved storing...

وصف كامل

محفوظ في:
التفاصيل البيبلوغرافية
المؤلف الرئيسي: Getnet Mezgebu Brhanemeskel (19517452) (author)
مؤلفون آخرون: Solomon Teferra Abate (19517455) (author), Tewodros Alemu Ayall (19273717) (author), Abegaz Mohammed Seid (19170901) (author)
منشور في: 2022
الموضوعات:
الوسوم: إضافة وسم
لا توجد وسوم, كن أول من يضع وسما على هذه التسجيلة!
_version_ 1864513506538160128
author Getnet Mezgebu Brhanemeskel (19517452)
author2 Solomon Teferra Abate (19517455)
Tewodros Alemu Ayall (19273717)
Abegaz Mohammed Seid (19170901)
author2_role author
author
author
author_facet Getnet Mezgebu Brhanemeskel (19517452)
Solomon Teferra Abate (19517455)
Tewodros Alemu Ayall (19273717)
Abegaz Mohammed Seid (19170901)
author_role author
dc.creator.none.fl_str_mv Getnet Mezgebu Brhanemeskel (19517452)
Solomon Teferra Abate (19517455)
Tewodros Alemu Ayall (19273717)
Abegaz Mohammed Seid (19170901)
dc.date.none.fl_str_mv 2022-11-18T09:00:00Z
dc.identifier.none.fl_str_mv 10.3390/app122211727
dc.relation.none.fl_str_mv https://figshare.com/articles/journal_contribution/Amharic_Speech_Search_Using_Text_Word_Query_Based_on_Automatic_Sentence-like_Segmentation/26889148
dc.rights.none.fl_str_mv CC BY 4.0
info:eu-repo/semantics/openAccess
dc.subject.none.fl_str_mv Information and computing sciences
Human-centred computing
Language, communication and culture
Language studies
Linguistics
speech segmentation
spoken term detection
automatic speech recognition
manual speech segmentation
dc.title.none.fl_str_mv Amharic Speech Search Using Text Word Query Based on Automatic Sentence-like Segmentation
dc.type.none.fl_str_mv Text
Journal contribution
info:eu-repo/semantics/publishedVersion
text
contribution to journal
description <p dir="ltr">More than 7000 languages are spoken in the world today. Amharic is one of the languages spoken in the East African country Ethiopia. A lot of speech data is being made every day in different languages as machines are getting better at processing and have improved storing capacity. However, searching for a particular word with its respective time frame inside a given audio file is a challenge. Since Amharic has its own distinguishing characteristics, such as glottal, palatal, and labialized consonants, it is not possible to directly use models that are developed for other languages. A popular approach in developing systems for searching particular information in speech involves using an automatic speech recognition (ASR) module that generates the text version of the speech where the word or phrase is searched based on text query. However, it is not possible to transcribe a long audio file without segmentation, which in turn affects the performance of the ASR module. In this paper, we are reporting our investigation on the effects of manual and automatic speech segmentation of Amharic audio files in a spiritual domain. We have used manual segmentation as a baseline for our investigation and found out that sentence-like automatic segmentation resulted in a word error rate (WER) closer to the WER achieved on the manually segmented test speech. Based on the experimental results, we propose Amharic speech search using text word query (ASSTWQ) based on automatic sentence-like segmentation. Since we have achieved lower WER using the previously developed speech corpus, which is in a broadcast news domain, together with the in-domain speech corpus, we recommend using both in- and out-domain speech corpora to develop the Amharic ASR module. The performance of the proposed ASR is a WER of 53% that needs further improvement. Combining two language models (LMs) developed using training text from the two different domains (spiritual and broadcast news) allowed a WER reduction from 53% to 46%. Therefore, we have developed two ASSTWQ systems using the two ASR modules with WERs of 53% and 46%.</p><h2>Other Information</h2><p dir="ltr">Published in: Applied Sciences<br>License: <a href="https://creativecommons.org/licenses/by/4.0/" target="_blank">https://creativecommons.org/licenses/by/4.0/</a><br>See article on publisher's website: <a href="https://dx.doi.org/10.3390/app122211727" target="_blank">https://dx.doi.org/10.3390/app122211727</a></p>
eu_rights_str_mv openAccess
id Manara2_f4f736a3f7e5b1fc6951136ec39c93fd
identifier_str_mv 10.3390/app122211727
network_acronym_str Manara2
network_name_str Manara2
oai_identifier_str oai:figshare.com:article/26889148
publishDate 2022
repository.mail.fl_str_mv
repository.name.fl_str_mv
repository_id_str
rights_invalid_str_mv CC BY 4.0
spelling Amharic Speech Search Using Text Word Query Based on Automatic Sentence-like SegmentationGetnet Mezgebu Brhanemeskel (19517452)Solomon Teferra Abate (19517455)Tewodros Alemu Ayall (19273717)Abegaz Mohammed Seid (19170901)Information and computing sciencesHuman-centred computingLanguage, communication and cultureLanguage studiesLinguisticsspeech segmentationspoken term detectionautomatic speech recognitionmanual speech segmentation<p dir="ltr">More than 7000 languages are spoken in the world today. Amharic is one of the languages spoken in the East African country Ethiopia. A lot of speech data is being made every day in different languages as machines are getting better at processing and have improved storing capacity. However, searching for a particular word with its respective time frame inside a given audio file is a challenge. Since Amharic has its own distinguishing characteristics, such as glottal, palatal, and labialized consonants, it is not possible to directly use models that are developed for other languages. A popular approach in developing systems for searching particular information in speech involves using an automatic speech recognition (ASR) module that generates the text version of the speech where the word or phrase is searched based on text query. However, it is not possible to transcribe a long audio file without segmentation, which in turn affects the performance of the ASR module. In this paper, we are reporting our investigation on the effects of manual and automatic speech segmentation of Amharic audio files in a spiritual domain. We have used manual segmentation as a baseline for our investigation and found out that sentence-like automatic segmentation resulted in a word error rate (WER) closer to the WER achieved on the manually segmented test speech. Based on the experimental results, we propose Amharic speech search using text word query (ASSTWQ) based on automatic sentence-like segmentation. Since we have achieved lower WER using the previously developed speech corpus, which is in a broadcast news domain, together with the in-domain speech corpus, we recommend using both in- and out-domain speech corpora to develop the Amharic ASR module. The performance of the proposed ASR is a WER of 53% that needs further improvement. Combining two language models (LMs) developed using training text from the two different domains (spiritual and broadcast news) allowed a WER reduction from 53% to 46%. Therefore, we have developed two ASSTWQ systems using the two ASR modules with WERs of 53% and 46%.</p><h2>Other Information</h2><p dir="ltr">Published in: Applied Sciences<br>License: <a href="https://creativecommons.org/licenses/by/4.0/" target="_blank">https://creativecommons.org/licenses/by/4.0/</a><br>See article on publisher's website: <a href="https://dx.doi.org/10.3390/app122211727" target="_blank">https://dx.doi.org/10.3390/app122211727</a></p>2022-11-18T09:00:00ZTextJournal contributioninfo:eu-repo/semantics/publishedVersiontextcontribution to journal10.3390/app122211727https://figshare.com/articles/journal_contribution/Amharic_Speech_Search_Using_Text_Word_Query_Based_on_Automatic_Sentence-like_Segmentation/26889148CC BY 4.0info:eu-repo/semantics/openAccessoai:figshare.com:article/268891482022-11-18T09:00:00Z
spellingShingle Amharic Speech Search Using Text Word Query Based on Automatic Sentence-like Segmentation
Getnet Mezgebu Brhanemeskel (19517452)
Information and computing sciences
Human-centred computing
Language, communication and culture
Language studies
Linguistics
speech segmentation
spoken term detection
automatic speech recognition
manual speech segmentation
status_str publishedVersion
title Amharic Speech Search Using Text Word Query Based on Automatic Sentence-like Segmentation
title_full Amharic Speech Search Using Text Word Query Based on Automatic Sentence-like Segmentation
title_fullStr Amharic Speech Search Using Text Word Query Based on Automatic Sentence-like Segmentation
title_full_unstemmed Amharic Speech Search Using Text Word Query Based on Automatic Sentence-like Segmentation
title_short Amharic Speech Search Using Text Word Query Based on Automatic Sentence-like Segmentation
title_sort Amharic Speech Search Using Text Word Query Based on Automatic Sentence-like Segmentation
topic Information and computing sciences
Human-centred computing
Language, communication and culture
Language studies
Linguistics
speech segmentation
spoken term detection
automatic speech recognition
manual speech segmentation