Hybrid Battery-friendly Mobile Solution for Extracting Users’ Visited Places

Extracting and analyzing outdoor humans’ activities represent a strong support for several applications fields, ranging from traffic management to marketing and social studies. Mobile users take their devices with them everywhere which leads to an increasing availability of persons’ traces used to recognize their activities. However, mobile environment is distinguished from one to another by its resources limitations. In this paper, we present a novel hybrid approach that combines activity recognition and prediction algorithms in order to online recognize users’ outdoor activities without draining the mobile resources. Our approach minimizes activity computations by wisely reducing the search frequency of activities, we demonstrate that our proposal is capable of reducing the battery consumption up to 60% while maintaining the same accuracy as its similar.