A small program to do simple analysis on facebook messenger data and list/graph outputs.
Runs in python3
https://www.facebook.com/help/1701730696756992?helpref=hc_global_nav
Note that you can download a lot more than just your messenger history, but this program only works on facebook messages.
Download your facebook messenger history as explained in the section above.
Unpack the history into the folder containing the analyze-messages.py file. Make sure you have the inbox directory the same level as the file, or you will need to manually set the path for the files.
Install the requirements using
pip3 install -r requirements.txt
Basic Run:
python3 analyze-messages.py <your facebook name>
There are also a number of options:
| Option | Long-form | Description |
|---|---|---|
| -h | --help | Display help message |
| -p | --path | Relative or absolute path to the location of the conversation folders. Default "./inbox" |
| -i | --include-threshold | The smallest number of total messages in a conversation for a conversation to be counted. Default 100 |
| -s | --sort-mode | How to sort the messages. Options: 'total', 'num_words', 'oldest', 'imgur'. Default "total" |
| -t | --top-people | The top number of people people to display. Default 5 |
| -f | --filter | List of names, separated by comma (no whitespace between), for the program to filter to. If not provided, will do no filtering. If a given name does not exist, will still filter, but will do nothing for that name |
| -so | --summary-only | Print only summary information in command-line. Do not display graphs |
| -ph | --print-history | Print the messaging history per month |
| -w | --word-count | Display analysis with word-count in addition to total messages |
| -b | --bar-mode | How to display the bars in the history graph. Options: 'group', 'stack'. Default "group" |
| -rg | --relative-graphs | Display the relative-percent graphs as well (kinda ugly if done with more than 3-4 people) |
| -c | --calls | Include graphs that display info about calls. Default False |
| -wc | --words-calls | Add call duration to word-count calculations based on 120wpm conversation discussion speed |
I've included two anonymized conversations as examples so that the program will run out of the box.
There are currently no tests for the program
I've included a script that can anonymize a messenger conversation. The script is anonymize.py. It takes three arguments: the file's location, the person running's name, and the other person's name.
It removes all instances of names from the file and lorem-ipsum-ifies the message contents while preserving the word-count.
Note that I've only run it on the two example conversations, so it might not fully anonymize all messages.
The script relies on plotly to generate its graphs. Please see plotly documentation for more info on plotly.