|
--- |
|
license: cc-by-nc-4.0 |
|
task_categories: |
|
- automatic-speech-recognition |
|
language: |
|
- th |
|
--- |
|
This dataset features natural Thai-language conversations between human speakers and machine agents, simulating real-world call center interactions across a variety of customer service domains. All dialogues are non-scripted and performed as role-play scenarios, capturing spontaneous, realistic exchanges. |
|
|
|
π£οΈ Speech Type: Human-to-machine conversations, simulating AI agents and human customers' dialogues. |
|
|
|
π Style: Spontaneous, unscripted role-playing, designed to reflect actual customer service situations. |
|
|
|
π§ Audio Format: High-quality stereo WAV recordings with separated speaker channels, captured in a quiet environment. |
|
|
|
π Transcriptions: Fully human-verified transcripts with 99%+ accuracy. |
|
|
|
π Metadata: Includes an accompanying metadata file detailing speaker attributes, topics, and recording conditions for each session. |
|
|
|
π License: This dataset is intended strictly for research and AI model fine-tuning. Commercial use, resale, or redistribution is prohibited. |
|
|
|
π Brought to you by AIxBlock β a decentralized platform for AI development and workflow automations, with a commitment to enabling the development of fair, accurate, and responsible AI systems through high-quality open datasets. |
|
|
|
Link to download full data (22GB): https://drive.google.com/drive/folders/1kxm9WI1temiXOUA-FQGhWFrQsMCSzNWC?usp=sharing |