Twitter Archive Analysis

A Python project for analyzing Twitter/X data archives.

Features

• Data Conversion: Converts Twitter's JavaScript data files to JSON format
• Ad Analysis: Analyzes ad impressions, targeting, and engagement data
• Visualizations: Creates charts showing targeting types, impressions over time, and more

Files

• convert-js.py - Converts Twitter export JS files to JSON
  • they are in a loosely wrapped javascript file originally for use with an html page included in the export.
• explore.ipynb - Jupyter notebook for data exploration and visualization
• export/ - Contains Twitter archive data (HTML viewer, JSON data, media)
  • this is ignored by git, the only file strictly required is ad-engagements.json nested under export/data/json but everything in json/ will be converted from js.

Usage

1. Install dependencies: uv sync
2. Run conversion script: uv run convert-js.py
3. Explore data in explore.ipynb. VSCode of Jupyter work here but make sure the notebook kernel is in the same created by uv.

Data Sources

Analyzes Twitter archive data including:

• Ad impressions and engagements
• Targeting criteria
• Tweet content and metadata
• Media files

Built with Python, pandas, matplotlib, and Jupyter.