Detecting social interactions via mobile sensing - Dataset

The goal of Markus' master thesis was to establish whether real-life social interactions could be detected using smartphone internal sensors.

The present dataset is a labelled set of sensor data. It contains sensing data of six persons, who each used the sensing application for three days in a row. During these days, the test persons in addition created labels for their real-life sensing data.

Participants in the first group in addition worked in the same office, and used the sensing at the same time. Their social interactions therefore overlap.