Mining for motivation: Using a single wearable accelerometer to detect people's interests

This paper presents a novel investigation of how motion as measured with just a single wearable accelerometer is informative of people's interests and motivation during crowded social events. We collected accelerometer readings on a large number of people (32 and 46 people in two crowded social events involving up to hundreds of people). In our experiments, we demonstrate how people's movements are informative of their particular interests: during talks, their interests in particular topics, and during networking events, their interest to participate successfully to make new contacts and foster existing ones. To our knowledge, using a single body worn accelerometer to measure and automatically infer these aspects of social behaviour has never been attempted before. Our experiments show that despite the challenge of the proposed task, useful automated predictions are possible and demonstrate the potential for further research in this area.