https://github.com/morriz/indy-news
Streamlit app and FastAPI that powers Indy News assistant
https://github.com/morriz/indy-news
ai bm25 independent-news media vector-database youtube
Last synced: 8 months ago
JSON representation
Streamlit app and FastAPI that powers Indy News assistant
- Host: GitHub
- URL: https://github.com/morriz/indy-news
- Owner: Morriz
- Created: 2023-12-07T07:13:30.000Z (almost 2 years ago)
- Default Branch: main
- Last Pushed: 2025-02-22T22:38:01.000Z (8 months ago)
- Last Synced: 2025-02-22T23:25:49.866Z (8 months ago)
- Topics: ai, bm25, independent-news, media, vector-database, youtube
- Language: Python
- Homepage: https://indy-news.streamlit.app/
- Size: 11.4 MB
- Stars: 2
- Watchers: 1
- Forks: 1
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Indy News
Try it on [Streamlit](https://indy-news.streamlit.app/)!
## Search for relevant independent media outlets or videos covering a topic
Contains both a streamlit app and a FastAPI module. The app uses the following _SMART_ api endpoints:
- [/media](http://127.0.0.1:8000/media?query=israel)
- [/youtube](http://127.0.0.1:8000/youtube?query=israel)
When I say smart I mean that the search tech used is:
- vector db
- hybrid bm25 + vector retriever
It also contains some dumb endpoints that only match input against the "name" property:
- [/allsides](http://127.0.0.1:8000/allsides?query=israel)
- [/mediabiasfactcheck](http://127.0.0.1:8000/mediabiasfactcheck?query=israel)
### Databases used (see `data/` folder)
- all.csv: 80+ media outlets enriched with entries from [mediabiasfactcheck.com](https://mediabiasfactcheck.com) data (if found).
- allsides.com.json: snapshot of the AllSides db
- mediabiasfactcheck.com.json: snapshot of the MediaBiasFactCheck db
### Original sources used
The following got normalized and ended up in all.csv:
- [libguides.rowan.edu](https://libguides.rowan.edu)
- [localfutures.org](https://localfutures.org)
- [trustworthymedia.org](https://trustworthymedia.org)
- some individuals and teams that I esteem and follow myself ;)
### Dev instructions
Run streamlit locally: `.venv/bin/streamlit run Home.py`
Run api locally: `.venv/bin/uvicorn --host "0.0.0.0" -p 8088`