An Approach to Modeling Call Response Behavior on Mobile Phones Based on Multi-Dimensional Contexts

Due to the popularity of context-aware computingand the rapid growth of the smart phone devices, modeling anindividual's phone call response behavior may assist them intheir daily activities for managing call interruptions. A key stepof such modeling is to discovering call response behavioral rulesbased on multi-dimensional contexts related to individual'sbehavior. Currently, researchers use classification rule learnersfor modeling individual's mobile phone behavior. However, theproblem is that such learning techniques produce only rulesthat include maximal number of contexts albeit ordered byrelevance. This results in many rules with low-reliability thatdecrease the accuracy of the modeling approach. In this paper, we propose an approach (Tmodel) to modeling individual'sphone call response behavior utilizing mobile phone data. Thisapproach produces not only general rules that capture individual'sbehavior at a particular level of confidence with a minimalnumber of contexts, but also produce rules that express specificexceptions to the general rules when more context-dimensionsare taken into account. Experimental evaluation shows thatour approach outperforms existing approaches to modelingindividual's phone call response behavior based on multidimensional contexts.