@inproceedings{abuzeina-etal-2012-rescoring,
title = "Rescoring N-Best Hypotheses for {A}rabic Speech Recognition: A Syntax- Mining Approach",
author = "AbuZeina, Dia and
Elshafei, Moustafa and
Al-Muhtaseb, Husni and
Al-Khatib, Wasfi",
editor = "Farghaly, Ali and
Oroumchian, Farhad",
booktitle = "Fourth Workshop on Computational Approaches to Arabic-Script-based Languages",
month = nov # " 1",
year = "2012",
address = "San Diego, California, USA",
publisher = "Association for Machine Translation in the Americas",
url = "https://aclanthology.org/2012.amta-caas14.8/",
pages = "57--64",
abstract = "Improving speech recognition accuracy through linguistic knowledge is a major research area in automatic speech recognition systems. In this paper, we present a syntax-mining approach to rescore N-Best hypotheses for Arabic speech recognition systems. The method depends on a machine learning tool (WEKA-3-6-5) to extract the N-Best syntactic rules of the Baseline tagged transcription corpus which was tagged using Stanford Arabic tagger. The proposed method was tested using the Baseline system that contains a pronunciation dictionary of 17,236 vocabularies (28,682 words and variants) from 7.57 hours pronunciation corpus of modern standard Arabic (MSA) broadcast news. Using Carnegie Mellon University (CMU) PocketSphinx speech recognition engine, the Baseline system achieved a Word Error Rate (WER) of 16.04 {\%} on a test set of 400 utterances ( about 0.57 hours) containing 3585 diacritized words. Even though there were enhancements in some tested files, we found that this method does not lead to significant enhancement (for Arabic). Based on this research work, we conclude this paper by introducing a new design for language models to account for longer-distance constrains, instead of a few proceeding words."
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="abuzeina-etal-2012-rescoring">
<titleInfo>
<title>Rescoring N-Best Hypotheses for Arabic Speech Recognition: A Syntax- Mining Approach</title>
</titleInfo>
<name type="personal">
<namePart type="given">Dia</namePart>
<namePart type="family">AbuZeina</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Moustafa</namePart>
<namePart type="family">Elshafei</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Husni</namePart>
<namePart type="family">Al-Muhtaseb</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Wasfi</namePart>
<namePart type="family">Al-Khatib</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2012-nov 1</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Fourth Workshop on Computational Approaches to Arabic-Script-based Languages</title>
</titleInfo>
<name type="personal">
<namePart type="given">Ali</namePart>
<namePart type="family">Farghaly</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Farhad</namePart>
<namePart type="family">Oroumchian</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Machine Translation in the Americas</publisher>
<place>
<placeTerm type="text">San Diego, California, USA</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Improving speech recognition accuracy through linguistic knowledge is a major research area in automatic speech recognition systems. In this paper, we present a syntax-mining approach to rescore N-Best hypotheses for Arabic speech recognition systems. The method depends on a machine learning tool (WEKA-3-6-5) to extract the N-Best syntactic rules of the Baseline tagged transcription corpus which was tagged using Stanford Arabic tagger. The proposed method was tested using the Baseline system that contains a pronunciation dictionary of 17,236 vocabularies (28,682 words and variants) from 7.57 hours pronunciation corpus of modern standard Arabic (MSA) broadcast news. Using Carnegie Mellon University (CMU) PocketSphinx speech recognition engine, the Baseline system achieved a Word Error Rate (WER) of 16.04 % on a test set of 400 utterances ( about 0.57 hours) containing 3585 diacritized words. Even though there were enhancements in some tested files, we found that this method does not lead to significant enhancement (for Arabic). Based on this research work, we conclude this paper by introducing a new design for language models to account for longer-distance constrains, instead of a few proceeding words.</abstract>
<identifier type="citekey">abuzeina-etal-2012-rescoring</identifier>
<location>
<url>https://aclanthology.org/2012.amta-caas14.8/</url>
</location>
<part>
<date>2012-nov 1</date>
<extent unit="page">
<start>57</start>
<end>64</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Rescoring N-Best Hypotheses for Arabic Speech Recognition: A Syntax- Mining Approach
%A AbuZeina, Dia
%A Elshafei, Moustafa
%A Al-Muhtaseb, Husni
%A Al-Khatib, Wasfi
%Y Farghaly, Ali
%Y Oroumchian, Farhad
%S Fourth Workshop on Computational Approaches to Arabic-Script-based Languages
%D 2012
%8 nov 1
%I Association for Machine Translation in the Americas
%C San Diego, California, USA
%F abuzeina-etal-2012-rescoring
%X Improving speech recognition accuracy through linguistic knowledge is a major research area in automatic speech recognition systems. In this paper, we present a syntax-mining approach to rescore N-Best hypotheses for Arabic speech recognition systems. The method depends on a machine learning tool (WEKA-3-6-5) to extract the N-Best syntactic rules of the Baseline tagged transcription corpus which was tagged using Stanford Arabic tagger. The proposed method was tested using the Baseline system that contains a pronunciation dictionary of 17,236 vocabularies (28,682 words and variants) from 7.57 hours pronunciation corpus of modern standard Arabic (MSA) broadcast news. Using Carnegie Mellon University (CMU) PocketSphinx speech recognition engine, the Baseline system achieved a Word Error Rate (WER) of 16.04 % on a test set of 400 utterances ( about 0.57 hours) containing 3585 diacritized words. Even though there were enhancements in some tested files, we found that this method does not lead to significant enhancement (for Arabic). Based on this research work, we conclude this paper by introducing a new design for language models to account for longer-distance constrains, instead of a few proceeding words.
%U https://aclanthology.org/2012.amta-caas14.8/
%P 57-64
Markdown (Informal)
[Rescoring N-Best Hypotheses for Arabic Speech Recognition: A Syntax- Mining Approach](https://aclanthology.org/2012.amta-caas14.8/) (AbuZeina et al., AMTA 2012)
ACL