jlumbroso/text-message-analysis-notebook

Jupyter notebook with examples on how to visualize the dataset of personal texts 📱, after extracting from an iPhone with PhoneView.

Text Message Analysis Notebook

This notebook shows how to visualize the number of text messages exchanged (with a specific person, or overall), and the volume of characters exchanged.

Rationale

It is surprising to notice how much information can be extracted from the visualization of the patterns of communication alone—without the contents of the messages. For instance, in the example above, it is clear that the person sending messages (in green) is more invested in the relationship: In the rescaled plot, we can see that in the beginning, both of the correspondents are responsible for about 50% of the messages; but this trend quickly changes over time.

Requirements

• PhoneView: This is the app I recommend to extract text messages from your iPhone. It is able to handle SMS, iMessage and WhatsApp seamlessly. It costs $29.95 for a Lifetime License, and has been available and reliable since 2008—but there is also a trial version that will work for this.

• Jupyter Notebook and pandas: These are the tools I use in this notebook to analyze and plot my text messages.

For Windows & Android users

For Windows computers, you can use iExplorer; for Android phones, you can use the SMS Backup and Restore app. Read this DigitalTrends article for more information. The notebook provided here may need to be adapted.