IEEE Themes - Learning about human behavior from mobile phone data

Themes introduction. Plus, Probalistic Mining Of Socio-Geographic Routines From Mobile Phone Data: Katayou Farrah presents this paper that suggests that human interaction data, or human proximity, obtained by Bluetooth sensor data, can be integrated with human location data, obtained by mobile cell tower connections, to mine meaningful details about human activities from large and noisy datasets. Farrah proposed a bag of multimodal behavior, which integrates the modeling of variations of location over multiple time-scales, and the modeling of interaction types from proximity. The representation is simple yet robust to characterize real-life human behavior sensed from mobile phones, which produce data known to be noisy or incomplete. Farrah describes an unsupervised approach, based on topic models, to discover latent human activities in terms of the joint interaction and location behaviors of 97 individuals over the course of a 10-month period. Some of the human activities discovered with a multimodal data representation include "going out alone from 7pm-midnight alone" and "working from 11am-5pm with 3-5 other people" occurring on Mondays. The paper also reports dominant work patterns occurring on other days of the week, such as Sunday work patterns, which occur from 5pm-midnight in small groups. The paper demonstrates the feasibility of our human routines discovered by predicting missing multimodal phone data, with a probabilistic topic model approach.