尚未偵測到更新。GitHub 發布版本後會顯示在這裡。
依最近已完成的 8 小時觀測回合間的 GitHub 星號增量排序
個人 GitHub 帳號下公開儲存庫的擁有者可以登入並驗證所有權，自行管理介紹，無需管理員核准。
登入並產生驗證檔案，在 30 分鐘內將其提交至此儲存庫的預設分支。其他分支或分叉儲存庫中的檔案無效。然後在下方檢查檔案，即可無需管理員核准直接管理專案。請保留檔案以維持管理權限。
如需登入，將前往登入頁面。登入後請回到此處，再次點擊所選的驗證按鈕。
在專案資料夾中執行 /nomorevibe verify，即可驗證所有權並自行管理介紹。
一分鐘完成驗證In crisis situations, humanitarian organizations rely on local, affected people for intelligence. Open Seldon leverages social media sentiment analysis of Telegram communities to democratize which voices are heard by transforming unstructured, multilingual public data into actionable humanitarian intelligence.
翻譯準備中
取自儲存庫 README
正在載入完整 README。 以下為已儲存的摘錄。
A visual analytics platform that transforms unstructured, multilingual Telegram data from refugee communities into actionable humanitarian intelligence.
• Scrapes public Telegram channels using the Telegram API
• Translates multilingual messages to English via Claude
• Classifies messages into a humanitarian needs taxonomy (IOM categories)
• Extracts and geocodes locations mentioned in messages
• Stores structured data and semantic embeddings in PostgreSQL + pgvector
• Serves an interactive dashboard: map, category distribution, streaming chatbot, executive summary generator檢視完整 README ↗A visual analytics platform that transforms unstructured, multilingual Telegram data from refugee communities into actionable humanitarian intelligence.
| Layer | Technology |
|---|---|
| Backend API | FastAPI (Python, async) |
| Database | PostgreSQL 16 + pgvector |
| AI/LLM | Claude API (Anthropic) |
| NER / categorization | GLiNER2 (HuggingFace) |
| Embeddings | all-MiniLM-L6-v2 (sentence-transformers) |
| Geocoding | Nominatim (geopy) |
| Frontend | Next.js (TypeScript) |
| Infrastructure | Docker Compose |
.env file (copy from .env.example and fill in values)cp .env.example .env
# fill in ANTHROPIC_API_KEY, TELEGRAM_API_ID, TELEGRAM_API_HASH, etc.
docker compose -f docker-compose.yml -f docker-compose.local.yml up --build
The API will be at http://localhost:8000, docs at http://localhost:8000/docs, and the web app at http://localhost:3000.
# Scrape a Telegram channel
docker compose exec api python -m pipeline.scraper --channel polska_grupa_informacyjna
# Scrape every enabled channel and process everything waiting (what the daily job runs)
docker compose exec api python -m pipeline.run_daily
api/ FastAPI backend + processing pipeline
web/ Next.js frontend
scripts/ Database init, utility scripts
.github/ CI workflows
See CODE_OF_CONDUCT.md. Issues and pull requests welcome.
MIT — see LICENSE.