ChatlogGrapher

Create pretty graphs of chat log history! Can combine information from multiple chat services all at once.

Supports:

• Facebook Messenger (JSON export)
• Discord (GDPR data request)
• Google Hangouts (Takeout)

How to use

• Download your data
  • Facebook Messenger
    • instructions at https://www.facebook.com/help/1701730696756992
    • Be sure to export in JSON format and include your chat data!
    • Unzip your download, make sure the folder is named something like "facebook-Name12345", and put it in data
  • Discord (GDPR request)
    • instructions at https://support.discordapp.com/hc/en-us/articles/360004027692-Requesting-a-Copy-of-your-Data
    • Unzip your download, rename the folder named "package" to something like "discord-YourUsername", and put it in data
    • NOTE that Discord data export only includes your sent messages, not received messages!!!!
  • Google Hangouts
    • instructions at https://support.google.com/accounts/answer/3024190?hl=en
    • Be sure to include your Hangouts (chat) data!
    • Unzip your download, rename the folder named "Takeout" to something like "google-YourUsername", and put it in data
• Install the dependencies
  • Make sure you have Conda installed
  • Set up the Conda environment: conda env create -f environment.yml
  • Activate the conda environment: conda activate chatloggrapher
  • If you don't want to have Conda installed... make sure you have the dependency packages installed: JupyterLab, numpy, pandas, matplotlib, gensim, nltk
• Run the code
  • Set up aliases.json to map multiple names/nicknames/usernames to each person's actual name
    • This is especially relevant for Discord, since Discord usernames are almost always different from someone's actual name and will also always include the 4-digit username ID.
    • If you have contacted the same person on multiple platforms where they may be listed under multiple names/usernames, make sure that those names are tied all to the same single name in the aliases file.
    • See the default lines in there for examples of how to write those lines.
    • Map any name/username to "null" to ignore it entirely.
  • Run jupyter lab and open the main notebook file data_processing.ipynb
  • Run the cells in order!

Other Info & Credits

Potential future features?

• Support for Facebook group chat data (> 2 person chats)
• Support for DiscordChatExporter data
• Support for any other chat or chat log scraper service that provides consistently formatted data dumps!
• Put the code that's currently in data_processing.ipynb into an independent Python file of its own, or something that can be more easily run!

Inspired by https://github.com/rohanp/MessengerGrapher (which handles messages through Facebook Messenger only)

Examples