oral-history

口述史录音转写稿 → 忠实整理稿的工具箱（Claude Skill）：去口语化 Pass1 + 对抗式忠实度质量门，存真优先，数据全程本地不外传

Web-based tool for correcting speech-to-text generated transcripts of oral histories.

This repository provides bulk download materials for potential collections as data projects, with materials adapted and extended from the Digital Library at the University of Utah

Oral History Text Analysis Project

A pipeline for topic modeling with oral history interviews

忆声 · 让人用说话的方式，留下一本还是本人在说的书。AI 采访者陪老人回忆人生，产出保真回忆录。

把英文书一键变成中文译本 + Obsidian 章节精读笔记 + 跨章主题索引。Claude Code skill。三产物独立可选：全本翻译 / 章节精读 / 主题归纳。

传家志 —— 会主动采访老人的 AI 记者小程序：家族口述记忆整理成编年史（微信小程序 + 云开发 + 阿里云百炼）

DigQDA is a modular AI-assisted methods lab for evidence-bound qualitative data analysis, with prompts and workflows for coding, interpretation, and human review.

Archive a spatial story in a single Markdown file — readers scroll it like an article, or step inside the scan. Local-first, no backend, hosts anywhere.

A memory vessel for the living — guided life interviews with a grounded LLM portrait and an honest witness. Built on Rusty.

Svelte 5 components that sync media playback with timed transcripts from IIIF manifests — for oral history, digital humanities, and time-based annotation.

Curated oral history and Holocaust testimony resources for AI/NLP, semantic search, ASR, RAG, and responsible AI.

Promotional page for the Oral History as Data (OHD) project

Openly available material, archive and lost-and-found: the complete recovered works of Asadullah Ali Al-Andalusi, founder of The Andalusian Project - 72 works, 20 papers, 68 videos, 68 transcripts - recovered from the Wayback Machine and his own surviving sites after the original web presence was lost, with per-item provenance.

Canonical engineering home, interactive 211-interview software craftsmanship oral history archive, 4D System Cartography case studies, and Principal Engineer resume.

Experimental oral-history transcript workflow with synthetic examples and human review.

DINOH Evaluation: synthetic examples and evaluation methods for oral-history assistance. Placeholder model backend; no model-performance results. OHPIPE is developed separately.

拾光·Echo：通过语音对话与证据门控五阶段流水线，将家族口述史整理成可分享、可续写的数字记忆卡。

Python pipeline that harvests qualitative research data from Harvard Dataverse & Columbia Oral History, catalogs QDA project files (.qdpx, NVivo, MAXQDA, ATLAS.ti) in SQLite, and classifies every project by ISIC Rev. 5 division using a stdlib TF-IDF classifier. Seeding QDArchive (SQ26) : FAU Erlangen-Nürnberg.