Composite Social Network for Predicting Mobile Apps Installation

摘要: We have carefully instrumented a large portion of the population living in university graduate dormitory by giving participants Android smart phones running our sensing software. In this paper, we propose novel problem predicting mobile application (known as "apps") installation using social networks and explain its challenge. Modern phones, like ones used study, are able to collect different built-in sensors. (e.g. Bluetooth proximity network, call log etc) While information is accessible app market makers such iPhone AppStore, it has not yet been studied how can use these for marketing research strategy development. develop simple computational model better predict composite network computed from sensed phones. Our also captures individual variance exogenous factors adoption. show importance considering all installations, observe surprising result that indeed predictable. achieves best results compared with generic approaches: four times than random guess, almost 45% apps users install precision (F1 score= 0.43).