Skip to content
Discover fun stats about your Facebook Messenger usage such as: most used words, most active year, most active hours, etc. using this Jupyter Notebook
Python Jupyter Notebook
Branch: master
Clone or download

README.md

Facebook Messenger Statistics

Have you downloaded your Facebook data? Are you curious/concerned about your Messenger usage? Discover fun stats about your Facebook Messenger usage such as: most used words, most active year, most active hours, etc.

Plots can be created using a Jupyter Notebook for exploratory analysis or a basic HTML report can be generated using the generate_report.py script.

demo

For instructions scroll down

Example Report

An interactive sample report can be found on my website report

How to use

Download your Messenger data from Facebook

Download your Facebook Messenger data in your settings page facebook settings

Make sure to download in JSON format. Media quality can be set to low for a faster download download

Clone this repository

Clone this repo using:

git clone git@github.com:simonwongwong/Facebook-Messenger-Statistics.git

And open the folder/navigate to the directory

Install Python and Python libraries

  1. If you don't have Python, install Python 3.7+
  2. Open a console on the project directory (or create your own custom environment) and run:
pip install -r "requirements.txt"

Generate an HTML report

Run the generate_report.py script either using your terminal or double-click the file on Windows

Locate your inbox folder from your extracted Facebook data using the file dialog. This folder should contain more folders for each chat and each of those folders should contain a message_1.json file. filedialog

If it runs successfully, you should receive a message similar to:

Parsing data from C:/Users/Simon/Desktop/messages/inbox
Report generated successfully!

And an HTML report should be at your current working directory.

Exploratory analysis on Jupyter

Open a console in the folder and start Jupyter Notebook using jupyter notebook command. From the Notebook file tree, open the Statistics notebook Statistics.ipynb

notebook

Make sure the correct directory is passed into loader.parse_from_json() and then have fun!

See chatstat.py or docstrings for parameters you can play with in the plot generators.

You can’t perform that action at this time.