de_quoi_parle_le_monde

Media Observer

A data / AI project to capture / analyse the evolution over time of the frontpages of main media sites.

A live version is available here : http://18.171.236.162:8000/

(Please forgive the ugly UI 🥹 ! You'd better use it on desktop, though there is an almost readable mobile version)

* Hosted on an AWS free-tier EC2 instance + "12 months free" RDS database, managed with Terraform

What is this ?

This project aims at observing what subjects news medias put forward on their websites.

The basic process consists of :

• finding snapshots of those sites as they were at precise times of the day (e.g. at 8h, 12h, 18h and 22h),
• parse those snapshots to extract relevant info (e.g. the main headline),
• store that info in a local database,
• find semantic similarities within the headlines using language models

A basic web UI is available to display the results.

At the moment, 6 sites are supported (see them there) but the list will expand over time.

None of this would be possible without the incredible Wayback Machine and the volunteers that have helped setup the snapshotting of all those sites for decades.

Installation

First you need to setup a PostgreSQL server and create a database whose path / credentials will be stored in a file .secrets.toml with the key database_url.

database_url="postgresql://user:password@yourdomain.com:port/database_name

With Rye

1. Install Rye project manager (following instructions from https://rye.astral.sh/guide/installation/)
2. Install dependencies : rye sync --no-lock --no-dev --all-features

Running the project

Setup your preferences by updating the configuration file

With Rye

• Do the site snapshots : rye run snapshots
• Compute the embeddings : rye run embeddings
• Build the similarity index : rye run similarity_index
• Run the web server : rye run web_server