Skip to content

Nexdata-AI/203-Hours-Korean-Medical-Entities-Real-world-Casual-Conversation-and-Monologue-speech-dataset

Repository files navigation

203-Hours-Korean-Medical-Entities-Real-world-Casual-Conversation-and-Monologue-speech-dataset


Description

Korean(Korea) Medical Entities Real-world Casual Conversation and Monologue speech dataset, covering various medical professional terminologies, primarily focuses on medical consultation, medical education, medical academic conferences and lectures, etc., mirrors real-world interactions. Transcribed with text content, speaker's ID, gender, common entities and other attributes. Our dataset was collected from extensive and diversify speakers, geographicly speaking, enhancing model performance in real and complex tasks. Quality tested by various AI companies. We strictly adhere to data protection regulations and privacy standards, ensuring the maintenance of user privacy and legal rights throughout the data collection, storage, and usage processes, our datasets are all GDPR, CCPA, PIPL complied.

For more details, please refer to the link: https://www.nexdata.ai/datasets/speechrecog/1499?source=Github

Specifications

Format

16k Hz, 16 bit, wav, mono channel

Content category

Covering various medical professional terminologies (such as diagnoses, medicines, etc.), primarily focuses on medical consultation, medical education, medical academic conferences and lectures, etc.

Recording condition

Low background noise

Country

Korea(KOR)

Language(Region) Code

ko-KR

Language

Korean

Features of annotation

transcription text, timestamp, speaker identification, gender, noise, PII redacted, entities, letter case

Accuracy

Word Accuracy Rate (WAR) at least 98%

Licensing Information

Commercial License

Releases

No releases published

Packages

No packages published