Some Issues in Modeling User Behavior Data in Systems of Engagement

The proliferation of mobile devices has changed the way digital information is consumed and its efficacy measured. These personal mobile devices know a lot about the user behavior from the sensors and activities performed by the user. This data can be used to provide a lot of personalized information to the user, and a measure of effectiveness of the information to the information providers. However, there are lot of challenges in modeling and storing such data from these systems of engagement so that there is a right balance of the redundancy in the data stored, and the usefulness of the data for analysis. In this paper we present an architecture and issues of modeling such user behavior data from different activities a user does on the mobile device while consuming some information. The user behavior data is modeled as NOSQL JSON documents and sent to a cloud backend where some simple MapReduce analytics can be done to understand different aspects of user preferences and information effectiveness.