Live Visualization and Editing of User Behavior in iObserve
Daniel Banck · 2017
The iObserve project collects user behavior information and derives from that information user behavior models. The visualization of these models supports operators to understand better how users are using an application, how performance bottlenecks can occur, and how the deployment of the system changes. In this thesis, we present an approach and implementation for a behavior model visualization service. The service provides a live visualization of user behavior graphs for operators to inspect. Furthermore, operators can create and modify user behavior graphs to reflect potential changes in user behavior and take future marketing campaigns into account. In an evaluation, we compare different Javascript libraries for their technical fit for the visualization and editor service.