Contextual Usage-Level Analysis of Mobile Services

Mobile services are used in a variety of contexts. The present paper studies how different services are used in different contexts. For this, the paper develops a specialized algorithm that can be used together with handset-based usage data acquired straight from handsets. With the algorithm a particular user's context can be divided into home, office and "on the move". The algorithm is further used with exemplary data from Finland and the UK. More than 70% of contextual use cases are correctly classified based on raw data available from handsets. Further improvement potential for the algorithm is discussed, and possible venues for future research identified. According to the exemplary results particularly multimedia services are used "on the move", whereas legacy mobile services experience more evenly distributed usage across all contexts. The combination of usage-level context identification and mobile end-user research with the presented platform holds lots of future potential.