SMS text analysis

This project is a follow-up to SMS Comics. During the 3 months trial of SMS Comics with Stanford students, we collected a large amount of SMS. This project was mostly designed to search for new findings in that collection of SMS, such as new group behaviors, the type of messages exchanged, etc.

I wrote the scripts to explore the dataset and extract statistics about the dataset, and created intermediate tools to manually categorize all text messages (Django server to randomly display a text message and let the user categorize).


A. Battestini, V. Setlur, T. Sohn. “A Large Scale Study of Text Messaging Use”, Mobile HCI, 2010