Scalable and Interactive Real-Time Visualization of Time Series Data

Time series data is everywhere. Everything that can be measured can be measured over time, which forms the definition of time series data. Visualization helps in perceiving this data by reducing the cognitive load. Interaction with the data can lead to an even better experience for users. This thesis evaluates an approach of an interactive and scalable visualization of time series data. The approach consists of enhancing an existing visualization library and adding new features to meet the requirements. The library already supports real-time visualization and the postulated support for user interactions. To enable its scalability, a cache to reduce bandwidth load is integrated. Moreover, prefetching algorithms increase the performance of the solution. After introducing the approach, this thesis also describes its integration into an existing web application. On this integration a feasibility and performance evaluation is conducted, which shows its advantages over the current version of the web application. Additionally, it proves the support for interactions like panning and zooming in the chart, as well as a real-time functionality. Moreover, the presented solution is scalable and does not show performance weaknesses in our test scenarios.