Do you like data? Graphs? Sentiment analysis? Have you waited the several-hours needed to download your Facebook Messenger chat history and wondered why you even bothered? Oh, do I ever have the Python scripts for you.
-
messages.pyprovides functions to load the chat history JSON file and count messages, words, emoji, stickers, photos, reacts, and links shared. Global totals are kept, as well as totals for each chat participant and every slice of time (length configurable). Sentiment (polarity and subjectivity) is analyzed for each message. All calculations / counts can be saved in JSON format.- Usage as command:
./messages.py [history_json_filename] [analysis_out_filename] [(a)ll|(y)ear|(m)onth|(w)eek|(d)ay]
- Usage as command:
-
plotstats.pyprovides functions to draw graphs based on the counts using matplotlib. It can draw stickers / images as x-axis labels, provided that their URIs as recorded bymessages.analyze()are valid files relative to the current working directory.- Usage as command:
./plotstats.py [analysis_filename]
- Usage as command:
depends on