Due to the increasing usage of eye tracking systems in the economical sector and academic research there is a rising demand for analyzation and visualization techniques for eye tracking data. This work applies the techniques and concepts of visual analytics by proposing a visualization-driven approach to analyze eye tracking data by filtering and analyzing the data before visualizing it. To do so, the data will be pre-processed. Afterwards the data can be analyzed by using configurable filters. The results are visualized using the Parallel Scan-Path visualization technique, enhanced by adding interaction methods such as brushing and zooming to enable further exploration of the data. The concept has been implemented as a prototype, which is evaluated in two scenarios.