Speech Data Collection Project for AI Korean Pronunciation

Job Description

Project Overview This project aims to **collect pronunciation and intonation data from non-native Korean speakers** to develop an AI system capable of accurately evaluating Korean conversational skills and providing **personalized learning paths**. The collected voice data will be used to **analyze pronunciation error patterns by nationality**, helping create **customized pronunciation correction and conversation training services** tailored to each learner’s native language and linguistic background. - Organizer: Korea Data Agency – “Data Voucher Support Program” (Selected Project) - Executor: **Langle** – AI-powered personalized Korean conversation learning platform - Data Usage: For pronunciation and intonation analysis and AI training purposes only. All data will be anonymized and **will not be used for any purpose other than AI training**. Participation Details - **Read 250–300 designated Korean sentences** (approx. 90–110 minutes including breaks) - Use a **high-quality microphone** in a **quiet, soundproof environment** - Includes orientation and rest periods - Re-recording may be required after quality review This project is conducted as part of the Ministry of Science and ICT’s Data Voucher Support Program. All collected voice data will be strictly anonymized and used only for AI training purposes.

Qualifications

Eligibility - Age 18 or older - Intermediate Korean proficiency or higher (able to read sentences fluently) - Applicants from the following countries/regions (initial recruitment focuses on English-speaking countries, with expansion planned later): - USA, UK, Canada, European countries, Korean diaspora living abroad - Balanced recruitment by nationality to ensure diversity in Korean pronunciation and intonation

Etc

Process 1. **Apply** via the Google Form (Name, Country, Korean proficiency level, Preferred date) 2. **Orientation** (Project introduction) 3. **Recording Session 1** (150 sentences) → Break 4. **Recording Session 2** (150 sentences) → Break 5. **Recording Session 3** (Re-recording, if required) → End 6. **Participation confirmation & reward payment**

Preferred Visas