Modular Bayesian Networks for Inferring Landmarks on Mobile Daily Life

Mobile devices get to handle much information thanks to the convergence of diverse functionalities. Their environment has great potential of supporting customized services to the users because it can observe the meaningful and private information continually for a long time. However, most of the information has been generally ignored because of the limitations of mobile devices. In this paper, we propose a novel method that infers landmarks efficiently in order to overcome the problems. It uses an effective probabilistic model of Bayesian networks for analyzing various log data on the mobile environment, which is modularized to decrease the complexity. The proposed methods are evaluated with synthetic mobile log data generated.