Open source smartphone libraries for computational social science

The ubiquity of sensor-rich and computationally powerful smartphones makes them an ideal platform for conducting social and behavioural research. However, building sensor data collection tools remains arduous and challenging: it requires an understanding of the varying sensor programming interfaces as well as the research issues related to building sensor-sampling systems. To alleviate this problem and facilitate the development of social sensing and data collection applications, we are developing a set of open-source smartphone libraries to collect, store and transfer, and query sensor data. Furthermore, we have also developed a library that can trigger notifications based on time or sensor events to assist experience sampling methods. This paper presents these libraries' architecture, initial feedback from developers using it, and a sensing application that we built using them to study daily affect.