面向国际中文教育的语音偏误数据集构建与应用

孙佳杰; 陈熙之; 凌锋; 袁丹; 兰韵诗; 王晔

doi:10.11959/j.issn.2096-0271.BDR26168

您当前的位置：

首页 >

文章列表页 >

面向国际中文教育的语音偏误数据集构建与应用

更新时间：2026-06-26

- 面向国际中文教育的语音偏误数据集构建与应用
- Construction and application of a pronunciation error dataset for international Chinese language education
- 大数据 2026年
- 作者机构：
  
  1.华东师范大学数据科学与工程学院，上海 200062
  2.华东师范大学国际汉语文化学院，上海 200062
- 作者简介：
  
  孙佳杰（2001-），男，硕士研究生，华东师范大学数据科学与工程学院，主要研究方向为智能教育。
  陈熙之（2000-），男，硕士，华东师范大学数据科学与工程学院，主要研究方向为大模型应用。
  凌锋（1976-），男，博士，华东师范大学国际汉语文化学院，副教授，主要研究方向为实验语音学、汉语方言学。
  袁丹（1982-），女，博士，华东师范大学国际汉语文化学院，副教授，主要研究方向为实验语音学、方言学、社会语言学以及二语习得。
  兰韵诗（1994-），女，博士，华东师范大学数据科学与工程学院，副教授，主要研究方向为知识图谱，智能问答以及其他与自然语言处理相关的任务。
  王晔（1977-），男，博士，华东师范大学数据科学与工程学院，专任研究员，主要研究方向为Web数据管理，海量数据挖掘，分布式系统。
- 基金信息：
- DOI：10.11959/j.issn.2096-0271.BDR26168
  中图分类号： TP391.41
- 收稿：2026-04-30，
  
  修回：2026-06-15，
  
  录用：2026-06-26，
- 稿件说明：
移动端阅览
孙佳杰, 陈熙之, 凌锋, 等. 面向国际中文教育的语音偏误数据集构建与应用[J/OL]. 大数据, 2026.

SUN Jiajie, CHEN Xizhi, LING Feng, et al. Construction and application of a pronunciation error dataset for international Chinese language education[J/OL]. Big Data Research, 2026.
孙佳杰, 陈熙之, 凌锋, 等. 面向国际中文教育的语音偏误数据集构建与应用[J/OL]. 大数据, 2026. DOI： 10.11959/j.issn.2096-0271.BDR26168.

SUN Jiajie, CHEN Xizhi, LING Feng, et al. Construction and application of a pronunciation error dataset for international Chinese language education[J/OL]. Big Data Research, 2026. DOI： 10.11959/j.issn.2096-0271.BDR26168.

摘要

针对国际中文教育中缺乏公开、标准化细粒度的汉语语音偏误数据集及通用模型非母语发音诊断适配有限的问题，构建面向留学生的汉语语音偏误数据集。采用融合基础声学对齐与专家诊断的十层结构化标注规范，建立“机器预标注—人工精修—专家复核”的人机协同流程，并完成数据清洗、质量复核与结构化入库，以保证样本可追溯和可训练。结果获得340份真实发音样本、12万余条声韵级时间对齐切片和数万条偏误诊断标签；在非母语口音自动语音识别和端到端语音偏误诊断任务中，基于该数据集微调的Paraformer-large与Qwen2-Audio-7B均表现出更好的任务适配效果。该数据集可为国际中文教育场景下的模型微调与智能发音反馈提供监督数据，也为垂直领域高质量数据集建设提供数据治理参考。

Abstract

To address the lack of open and standardized fine-grained pronunciation error datasets for international Chinese language education and the limited adaptation of general models to non-native pronunciation diagnosis

a Chinese pronunciation error dataset for international students was constructed. A ten-layer structured annotation scheme was developed by integrating basic acoustic alignment with expert diagnostic labels

and a human-machine collaborative workflow of machine pre-annotation

manual refinement

and expert review was established. Data cleaning

quality review

and structured storage were also completed to ensure that the samples were traceable and trainable. As a result

340 authentic speech samples

more than 120

000 time-aligned slices at the initial-final level

and tens of thousands of fine-grained diagnostic labels were obtained. To evaluate the dataset

two representative tasks were conducted: automatic speech recognition for non-native accented speech and end-to-end pronunciation error diagnosis. The results showed that

after fine-tuning on the dataset

Paraformer-large and Qwen2-Audio-7B achieved better task adaptation. The dataset provides supervised data for model fine-tuning and intelligent pronunciation feedback in international Chinese language education

and it also offers a reference for high-quality dataset governance in vertical domains.

关键词

Keywords

references

吴应辉 . 国际中文教育发展的新定位——基于2024世界中文大会领导人讲话的思考 [J ] . 昆明学院学报 , 2025 , 47 ( 1 ): 1 - 8 . DOI: 10.14091/j.cnki.kmxyxb.2025.01.001.Wu Y H. New Positioning for the Development of International Chinese Language Education-Reflections Based on the Speeches of Leaders at the 2024 World Chinese Language Conference http://dx.doi.org/10.14091/j.cnki.kmxyxb.2025.01.001.WuYH.NewPositioningfortheDevelopmentofInternationalChineseLanguageEducation-ReflectionsBasedontheSpeechesofLeadersatthe2024WorldChineseLanguageConference [J ] . Journal of Kunming University, 2025, 47 ( 1 ): 1 - 8 .

习近平向2024世界中文大会致贺信 [N ] . 人民日报 , 2024-11-16 (02).

Xi Jinping Sends a Congratulatory Letter to the 2024 World Chinese Language Conference [N ] . People's Daily , 2024-11-16 (02).

JIAO B . Phonetic Learning Strategies for Chinese as a Second Language [J ] . Journal of Education and Educational Research , 2024 , 8 ( 2 ): 199 - 201 . DOI: 10.54097/851k6918 http://dx.doi.org/10.54097/851k6918 .

BU H , DU J , NA X , et al . AISHELL-1: An open-source Mandarin speech corpus and a speech recognition baseline [C ] // 2017 20th Conference of the Oriental Chapter of the International Coordinating Committee on Speech Databases and Speech I/O Systems and Assessment (O-COCOSDA) . Seoul, South Korea : IEEE , 2017 : 1 - 5 . DOI: 10.1109/ICSDA.2017.8384449 http://dx.doi.org/10.1109/ICSDA.2017.8384449 .

ZHANG J , ZHANG Z , WANG Y , et al . speechocean762: An Open-Source Non-native English Speech Corpus For Pronunciation Assessment [C ] // Proceedings of Interspeech 2021 . Brno, Czech Republic : ISCA , 2021 : 3710 – 3714 . DOI: 10.21437/Interspeech.2021-1259 http://dx.doi.org/10.21437/Interspeech.2021-1259 .

WANG W , ZHANG J . Factors predicting human performance in error annotation for non-native speech corpus [J ] . Speech Communication , 2023 , 149 : 38 - 46 . DOI: 10.1016/j.specom.2023.03.001 http://dx.doi.org/10.1016/j.specom.2023.03.001 .

FANG Y , PENG J , LI X , et al . Low-Resource Domain Adaptation for Speech LLMs via Text-Only Fine-Tuning [A/OL ] . arXiv , 2025 . https://arxiv.org/abs/2506.05671 https://arxiv.org/abs/2506.05671 . DOI: 10.48550/ARXIV.2506.05671 http://dx.doi.org/10.48550/ARXIV.2506.05671 .

PENG L , GAO Y , BAO R , et al . End-to-End Mispronunciation Detection and Diagnosis Using Transfer Learning [J ] . Applied Sciences , 2023 , 13 ( 11 ): 6793 . DOI: 10.3390/app13116793 http://dx.doi.org/10.3390/app13116793 .

CAO W , WANG D , ZHANG J , et al . Developing a Chinese L2 speech database of Japanese learners with narrow-phonetic labels for computer assisted pronunciation training [C ] // Proceedings of Interspeech 2010 . Makuhari, Chiba, Japan : ISCA , 2010 : 1922 - 1925 . DOI: 10.21437/Interspeech.2010-553 http://dx.doi.org/10.21437/Interspeech.2010-553 .

SYLOLYPAVAN A , SLEEMAN D , WU H , et al . The impact of inconsistent human annotations on AI driven clinical decision making [J ] . npj Digital Medicine , 2023 , 6 : 26 . DOI: 10.1038/s41746-023-00773-3 http://dx.doi.org/10.1038/s41746-023-00773-3 .

RADFORD A , KIM J W , XU T , et al . Robust speech recognition via large-scale weak supervision [C ] // Proceedings of the 40th International Conference on Machine Learning . Honolulu, Hawaii, USA : PMLR , 2023 : 28492 - 28518 .

ALRASHOUDI N , AL-KHALIFA H , ALOTAIBI Y . Improving mispronunciation detection and diagnosis for non-native learners of the Arabic language [J ] . Discover Computing , 2025 , 28, Article 1 . DOI: 10.1007/s10791-024-09489-8 http://dx.doi.org/10.1007/s10791-024-09489-8 .

LIN Y X , YANG C K , CHEN W C , et al . A Preliminary Exploration with GPT-4o Voice Mode [A/OL ] . arXiv , 2025 [ 2026-06-09 ] . DOI: 10.48550/arXiv.2502.09940 http://dx.doi.org/10.48550/arXiv.2502.09940 .

FANG Y H , LO T H , SUNG Y T , et al . Beyond Modality Limitations: A Unified MLLM Approach to Automated Speaking Assessment with Effective Curriculum Learning [C ] // 2025 IEEE Automatic Speech Recognition and Understanding Workshop (ASRU) . Honolulu, Hawaii, USA : IEEE , 2025 . DOI: 10.48550/arXiv.2508.12591 http://dx.doi.org/10.48550/arXiv.2508.12591 .

MCAULIFFE M , SOCOLOF M , MIHUC S , et al . Montreal Forced Aligner: Trainable text-speech alignment using Kaldi [C ] // Proceedings of Interspeech 2017 . Stockholm, Sweden : ISCA , 2017 : 498 - 502 . DOI: 10.21437/Interspeech.2017-1386 http://dx.doi.org/10.21437/Interspeech.2017-1386 .

WILLIAMS S , FOULKES P , HUGHES V . Analysis of forced aligner performance on L2 English speech [J ] . Speech Communication , 2024 , 158 : 103042 . DOI: 10.1016/j.specom.2024.103042 http://dx.doi.org/10.1016/j.specom.2024.103042 .

PARIKH A K , TEJEDOR-GARCIA C , CUCCHIARINI C , et al . Enhancing GOP in CTC-Based Mispronunciation Detection with Phonological Knowledge [C ] // Proceedings of Interspeech 2025 . Rotterdam, Netherlands : ISCA , 2025 : 5068 - 5072 . DOI: 10.21437/Interspeech.2025-829 http://dx.doi.org/10.21437/Interspeech.2025-829 .

KORZEKWA D , LORENZO-TRUEBA J , ZAPOROWSKI S , et al . Mispronunciation Detection in Non-native (L2) English with Uncertainty Modeling [A/OL ] . arXiv , 2021 [ 2026-06-09 ] . DOI: 10.48550/arXiv.2101.06396 http://dx.doi.org/10.48550/arXiv.2101.06396 .

WITT S M , YOUNG S J . Phone-level pronunciation scoring and assessment for interactive language learning [J ] . Speech Communication , 2000 , 30 ( 2-3 ): 95 - 108 . DOI: 10.1016/S0167-6393(99)00044-8 http://dx.doi.org/10.1016/S0167-6393(99)00044-8 .

FENG Y , FU G , CHEN Q , et al . SED-MDD: Towards Sentence Dependent End-To-End Mispronunciation Detection and Diagnosis [C/OL ] // ICASSP 2020 - 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) . IEEE , 2020 : 3492 - 3496 . http://dx.doi.org/10.1109/icassp40776.2020.9052975 http://dx.doi.org/10.1109/icassp40776.2020.9052975 . DOI: 10.1109/ICASSP40776.2020.9052975 http://dx.doi.org/10.1109/ICASSP40776.2020.9052975 .

HU E J , SHEN Y , WALLIS P , et al . LoRA: low-rank adaptation of large language models [C/OL ] // International Conference on Learning Representations . 2022[2026-06-10] . https://openreview.net/forum?id=nZeVKeeFYf9 https://openreview.net/forum?id=nZeVKeeFYf9 .

ZHU C , WUMAIER A , WEI D , et al . Pronunciation error detection model based on feature fusion [J ] . Speech Communication , 2024 , 156 : 103009 . DOI: 10.1016/j.specom.2023.103009 http://dx.doi.org/10.1016/j.specom.2023.103009 .

RABINER L R , SAMBUR M R . An algorithm for determining the endpoints of isolated utterances [J ] . The Bell System Technical Journal , 1975 , 54 ( 2 ): 297 - 315 . DOI: 10.1002/j.1538-7305.1975.tb02840.x http://dx.doi.org/10.1002/j.1538-7305.1975.tb02840.x .

SAINBURG T , ZOREA A . Domain general noise reduction for time series signals with Noisereduce [J ] . Scientific Reports , 2025 , 15 : 30905 . DOI: 10.1038/s41598-025-13108-x http://dx.doi.org/10.1038/s41598-025-13108-x .

JADOUL Y , THOMPSON B , DE BOER B . Introducing Parselmouth: A Python interface to Praat [J ] . Journal of Phonetics , 2018 , 71 : 1 - 15 . DOI: 10.1016/j.wocn.2018.07.001 http://dx.doi.org/10.1016/j.wocn.2018.07.001 .

中华人民共和国教育部 . 2018年来华留学统计 [EB/OL ] . 2019-04-12 [ 2026-06-10 ] . https://www.moe.gov.cn/jyb_xwfb/gzdt_gzdt/s5987/201904/t20190412_377692.html.Ministry of Education of the People https://www.moe.gov.cn/jyb_xwfb/gzdt_gzdt/s5987/201904/t20190412_377692.html.MinistryofEducationofthePeople

's Republic of China . Statistics on international students in China in 2018 [EB/OL ] . 2019-04-12 [ 2026-06-10 ] . https://www.moe.gov.cn/jyb_xwfb/gzdt_gzdt/s5987/201904/t20190412_377692.html https://www.moe.gov.cn/jyb_xwfb/gzdt_gzdt/s5987/201904/t20190412_377692.html .

中华人民共和国教育部 . 教外〔2018〕50号. 教育部关于印发《来华留学生高等教育质量规范（试行）》的通知 [S/OL ] . 2018-09-03 [ 2026-06-10 ] .

Ministry of Education of the People's Republic of China . Document No. 50 [2018]. Notice of the Ministry of Education on issuing the Quality Standards for Higher Education for International Students in China (Trial) [S/OL ] . 2018-09-03 [ 2026-06-10 ] . https://www.moe.gov.cn/srcsite/A20/moe_850/201810/t20181012_351302.html.

https://www.moe.gov.cn/srcsite/A20/moe_850/201810/t20181012_351302.html https://www.moe.gov.cn/srcsite/A20/moe_850/201810/t20181012_351302.html .

曾志平 , 张旭龙 , 瞿晓阳 , 等 . 深度伪造音频生成与鉴伪技术综述 [J ] . 大数据 , 2025 , 11 ( 5 ): 130 - 151 . DOI: 10.11959/j.issn.2096-0271.2025064.Zeng Z P http://dx.doi.org/10.11959/j.issn.2096-0271.2025064.ZengZP ,

Zhang X L , Qu X Y , et al . Survey of deep fake audio generation and detection techniques [J ] . BIG DATA RESEARCH , 2025 , 11 ( 5 ): 130 - 151 .

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

面向高质量企业征信数据集的治理体系构建与实践

基于湖仓一体技术的“多湖一中台”工业大数据平台研究与应用实践

政务数据标识技术研究进展及下一代政务数据标识体系

决策驱动型省级政务大数据治理技术架构及实现

数据中台框架与实践