Khan Academy Turkish Audio Clips for Speech Recognition

78 hours of audio extracted from the Khan Academy Turkish YouTube channel. The dataset is segmented into short clips averaging 10.5 seconds each and was created by author ysdede, with a last recorded update on 2025-02 11.

Use Cases

• Train Turkish speech recognition models based on educational audio content.
• Develop text-to-speech systems for educational STEM topics.
• Fine-tune audio processing algorithms on segmented, short-duration clips.
• Analyze speech patterns in Turkish educational lectures.

Strengths

• Contains 78 hours of audio data.
• Clips are pre-segmented to an average duration of 10.5 seconds.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• Description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

Khan Academy Turkish YouTube channel.

Collection Method

Audio was extracted from YouTube and processed into short clips.

Time Range

null

Freshness

Last updated 2025-02-11 19:42:41; freshness should be verified.

Geography

null