The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. ex. Some numerals are expressed as "XNUMX".
Copyrights notice
The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. Copyrights notice
Kesilapan pengecaman kata nama khas dan perkataan asing dengan ketara mengurangkan prestasi aplikasi pertuturan berasaskan ASR seperti sistem dailan suara, ringkasan pertuturan, perolehan semula dokumen pertuturan dan perolehan semula maklumat berasaskan pertanyaan (IR). Sebabnya ialah kata nama khas dan perkataan yang berasal daripada bahasa lain biasanya merupakan kata kunci yang paling penting. Kehilangan kata-kata tersebut akibat salah pengecaman seterusnya membawa kepada kehilangan maklumat penting daripada sumber pertuturan. Kertas kerja ini memfokuskan kepada cara meningkatkan prestasi ASR Indonesia dengan mengurangkan masalah variasi sebutan kata nama khas dan perkataan asing (khususnya perkataan Inggeris). Untuk mempertingkatkan ketepatan pengecaman kata nama yang betul, model akustik khusus kata nama khas dicipta melalui penyesuaian diselia menggunakan regresi linear kemungkinan maksimum (MLLR). Untuk meningkatkan pengecaman perkataan Inggeris, sebutan perkataan Inggeris yang terkandung dalam leksikon ditetapkan dengan menggunakan pemetaan fonem Inggeris-ke-Indonesia berasaskan peraturan. Keberkesanan kaedah yang dicadangkan telah disahkan melalui pertanyaan lisan berasaskan IR Indonesia. Kami menggunakan IR berasaskan Rangkaian Inferens (berasaskan IN) dan membandingkan keputusannya dengan IR Model Ruang Vektor (VSM) klasik, kedua-duanya menggunakan skema pemberat tf-idf. Keputusan eksperimen menunjukkan bahawa IR berasaskan IN mengatasi IR VSM.
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Salinan
Dessi Puji LESTARI, Sadaoki FURUI, "Adaptation to Pronunciation Variations in Indonesian Spoken Query-Based Information Retrieval" in IEICE TRANSACTIONS on Information,
vol. E93-D, no. 9, pp. 2388-2396, September 2010, doi: 10.1587/transinf.E93.D.2388.
Abstract: Recognition errors of proper nouns and foreign words significantly decrease the performance of ASR-based speech applications such as voice dialing systems, speech summarization, spoken document retrieval, and spoken query-based information retrieval (IR). The reason is that proper nouns and words that come from other languages are usually the most important key words. The loss of such words due to misrecognition in turn leads to a loss of significant information from the speech source. This paper focuses on how to improve the performance of Indonesian ASR by alleviating the problem of pronunciation variation of proper nouns and foreign words (English words in particular). To improve the proper noun recognition accuracy, proper-noun specific acoustic models are created by supervised adaptation using maximum likelihood linear regression (MLLR). To improve English word recognition, the pronunciation of English words contained in the lexicon is fixed by using rule-based English-to-Indonesian phoneme mapping. The effectiveness of the proposed method was confirmed through spoken query based Indonesian IR. We used Inference Network-based (IN-based) IR and compared its results with those of the classical Vector Space Model (VSM) IR, both using a tf-idf weighting schema. Experimental results show that IN-based IR outperforms VSM IR.
URL: https://global.ieice.org/en_transactions/information/10.1587/transinf.E93.D.2388/_p
Salinan
@ARTICLE{e93-d_9_2388,
author={Dessi Puji LESTARI, Sadaoki FURUI, },
journal={IEICE TRANSACTIONS on Information},
title={Adaptation to Pronunciation Variations in Indonesian Spoken Query-Based Information Retrieval},
year={2010},
volume={E93-D},
number={9},
pages={2388-2396},
abstract={Recognition errors of proper nouns and foreign words significantly decrease the performance of ASR-based speech applications such as voice dialing systems, speech summarization, spoken document retrieval, and spoken query-based information retrieval (IR). The reason is that proper nouns and words that come from other languages are usually the most important key words. The loss of such words due to misrecognition in turn leads to a loss of significant information from the speech source. This paper focuses on how to improve the performance of Indonesian ASR by alleviating the problem of pronunciation variation of proper nouns and foreign words (English words in particular). To improve the proper noun recognition accuracy, proper-noun specific acoustic models are created by supervised adaptation using maximum likelihood linear regression (MLLR). To improve English word recognition, the pronunciation of English words contained in the lexicon is fixed by using rule-based English-to-Indonesian phoneme mapping. The effectiveness of the proposed method was confirmed through spoken query based Indonesian IR. We used Inference Network-based (IN-based) IR and compared its results with those of the classical Vector Space Model (VSM) IR, both using a tf-idf weighting schema. Experimental results show that IN-based IR outperforms VSM IR.},
keywords={},
doi={10.1587/transinf.E93.D.2388},
ISSN={1745-1361},
month={September},}
Salinan
TY - JOUR
TI - Adaptation to Pronunciation Variations in Indonesian Spoken Query-Based Information Retrieval
T2 - IEICE TRANSACTIONS on Information
SP - 2388
EP - 2396
AU - Dessi Puji LESTARI
AU - Sadaoki FURUI
PY - 2010
DO - 10.1587/transinf.E93.D.2388
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E93-D
IS - 9
JA - IEICE TRANSACTIONS on Information
Y1 - September 2010
AB - Recognition errors of proper nouns and foreign words significantly decrease the performance of ASR-based speech applications such as voice dialing systems, speech summarization, spoken document retrieval, and spoken query-based information retrieval (IR). The reason is that proper nouns and words that come from other languages are usually the most important key words. The loss of such words due to misrecognition in turn leads to a loss of significant information from the speech source. This paper focuses on how to improve the performance of Indonesian ASR by alleviating the problem of pronunciation variation of proper nouns and foreign words (English words in particular). To improve the proper noun recognition accuracy, proper-noun specific acoustic models are created by supervised adaptation using maximum likelihood linear regression (MLLR). To improve English word recognition, the pronunciation of English words contained in the lexicon is fixed by using rule-based English-to-Indonesian phoneme mapping. The effectiveness of the proposed method was confirmed through spoken query based Indonesian IR. We used Inference Network-based (IN-based) IR and compared its results with those of the classical Vector Space Model (VSM) IR, both using a tf-idf weighting schema. Experimental results show that IN-based IR outperforms VSM IR.
ER -