---
dataset_info:
  features:
    - name: video_id
      dtype: string
    - name: title
      dtype: string
    - name: source
      dtype: string
    - name: url
      dtype: string
    - name: duration_seconds
      dtype: int64
    - name: status
      dtype: string
    - name: priority
      dtype: int64
license: mit
task_categories:
  - automatic-speech-recognition
language:
  - en
  - ru
  - hi
  - es
  - pt
  - de
  - fr
  - ja
  - ko
  - zh
  - ar
  - tr
  - pl
tags:
  - education
  - youtube
  - video-ids
  - queue
size_categories:
  - 1M<n<10M
---

# Massive YouTube Educational Video Queue

**1M+ curated educational YouTube video IDs** with metadata, quality-filtered and priority-scored. Use this to build your own transcription pipeline or educational content index.

## Stats

| Metric | Value |
|--------|-------|
| Total videos | 1,014,726 |
| Pending transcription | ~920K |
| Completed (transcripts in [sibling dataset](https://huggingface.co/datasets/thepowerfuldeez/massive-yt-edu-transcriptions)) | ~7K |
| Languages | 15+ |

## Priority System

| Priority | Description | Count |
|----------|-------------|-------|
| P9 | University courses (MIT OCW, Stanford, NPTEL), academic conferences (NeurIPS, ICML) | ~20K |
| P8 | Lectures, tutorials, known edu creators (3Blue1Brown, Khan Academy) | ~200K |
| P7 | Documentaries, explainers, conference talks | ~5K |
| P5 | General educational content passing quality filter | ~790K |

## Quality Filtering Applied

- **Duration gate**: All videos ≥15 minutes (deep educational content only)
- **40+ reject categories**: Gaming, music videos, ASMR, vlogs, drama, pranks, mukbang, clickbait, conspiracy, sports highlights, etc.
- **Deduplication**: UNIQUE constraint on video_id

## Discovery Sources

- YouTube search saturation (subject × language × modifier)
- 140+ educational channel crawls
- Related video chaining (exponential discovery)
- Playlist walking
- Multi-language queries (15+ languages)

## Schema

```python
{
    "video_id": "dQw4w9WgXcQ",
    "title": "MIT 6.006 Lecture 1: Algorithms and Computation",
    "source": "MIT OpenCourseWare",
    "url": "https://youtube.com/watch?v=dQw4w9WgXcQ",
    "duration_seconds": 3600,
    "status": "pending",       # pending/completed/processing/error/rejected
    "priority": 9
}
```

## Usage

```python
from datasets import load_dataset

ds = load_dataset("thepowerfuldeez/massive-yt-edu-queue")

# Get high-priority university content
university = ds["train"].filter(lambda x: x["priority"] >= 9)

# Get pending videos for your own transcription
pending = ds["train"].filter(lambda x: x["status"] == "pending")
```

## Related

- **Transcriptions**: [thepowerfuldeez/massive-yt-edu-transcriptions](https://huggingface.co/datasets/thepowerfuldeez/massive-yt-edu-transcriptions) — completed transcripts (updated daily)
- **Source code**: [github.com/thepowerfuldeez/massive_yt_edu_scraper](https://github.com/thepowerfuldeez/massive_yt_edu_scraper)

## License

MIT. Video metadata is derived from publicly available YouTube data.