ishenwei/nexus
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Update nexus: fix conflicts and sync local changes

Browse Source
This commit is contained in:
Shen Wei
2026-04-26 12:06:50 +08:00
parent 191797c01b
commit f09834b5a5
2443 changed files with 254323 additions and 255154 deletions
Download Patch File Download Diff File Expand all files Collapse all files

76
wiki/concepts/Transcript-Based-Summarization.md
View File

@@ -1,38 +1,38 @@
---
title: "Transcript-Based Summarization"
type: concept
tags: [Transcript, Summarization, YouTube, Content-Processing, AI]
sources: [daily-youtube-digest]
last_updated: 2026-04-22
---
## Definition
Transcript-Based Summarization 是指从视频/音频内容中提取字幕/ transcript，然后通过 AI 压缩为结构化要点摘要的处理流程。它使长视频/播客的消费从"没时间看完"变为"5 分钟掌握精华"。
## Workflow
1. **Transcript Extraction**: 通过 API（[[TranscriptAPI.com]]）或 CLI 工具（yt-dlp）获取字幕
2. **AI Summarization**: LLM 处理字幕文本，输出关键点、亮点引用、时间戳
3. **Structured Output**: 生成 bullet points、key quotes、timestamps 等结构化格式
4. **Delivery**: 整合到 [[Daily-Digest]] 或 [[second-brain]]
## TranscriptAPI vs yt-dlp
| Criteria | yt-dlp | TranscriptAPI.com |
|---|---|---|
| Output format | Verbose CLI logs | Clean JSON |
| Cloud compatibility | Fails on GCP/cloud | ✅ Works everywhere |
| Caching | None | ✅ Cached results |
| Rate limiting | Random blocks | ✅ Reliable, millions served |
| Dependencies | Binary required | HTTP API only |
## Applications
- [[Daily YouTube Digest]]: 频道新视频 → 字幕提取 → 要点摘要 → 推送
- [[Podcast Production Pipeline]]: 播客音频 → 字幕 → 时间戳笔记 → 社交媒体片段
- [[youtube-content-pipeline]]: YouTube 视频 → 字幕 → 博客文章/Newsletter
## Connections
- [[Transcript-Based Summarization]] ← uses ← [[TranscriptAPI.com]]
- [[Daily-Digest]] ← incorporates ← [[Transcript-Based Summarization]]
---
title: "Transcript-Based Summarization"
type: concept
tags: [Transcript, Summarization, YouTube, Content-Processing, AI]
sources: [daily-youtube-digest]
last_updated: 2026-04-22
---
## Definition
Transcript-Based Summarization 是指从视频/音频内容中提取字幕/ transcript，然后通过 AI 压缩为结构化要点摘要的处理流程。它使长视频/播客的消费从"没时间看完"变为"5 分钟掌握精华"。
## Workflow
1. **Transcript Extraction**: 通过 API（[[TranscriptAPI.com]]）或 CLI 工具（yt-dlp）获取字幕
2. **AI Summarization**: LLM 处理字幕文本，输出关键点、亮点引用、时间戳
3. **Structured Output**: 生成 bullet points、key quotes、timestamps 等结构化格式
4. **Delivery**: 整合到 [[Daily-Digest]] 或 [[second-brain]]
## TranscriptAPI vs yt-dlp
| Criteria | yt-dlp | TranscriptAPI.com |
|---|---|---|
| Output format | Verbose CLI logs | Clean JSON |
| Cloud compatibility | Fails on GCP/cloud | ✅ Works everywhere |
| Caching | None | ✅ Cached results |
| Rate limiting | Random blocks | ✅ Reliable, millions served |
| Dependencies | Binary required | HTTP API only |
## Applications
- [[Daily YouTube Digest]]: 频道新视频 → 字幕提取 → 要点摘要 → 推送
- [[Podcast Production Pipeline]]: 播客音频 → 字幕 → 时间戳笔记 → 社交媒体片段
- [[youtube-content-pipeline]]: YouTube 视频 → 字幕 → 博客文章/Newsletter
## Connections
- [[Transcript-Based Summarization]] ← uses ← [[TranscriptAPI.com]]
- [[Daily-Digest]] ← incorporates ← [[Transcript-Based Summarization]]