Event abstraction is a technique that involves simplifying complex process models by grouping similar events together and representing them using abstracted entities. This technique can be effectively utilized to improve the readability and manageability of complex process models derived from event logs in the following ways:

1. Reducing model complexity: Event abstraction can help reduce the complexity of a process model by grouping similar events into higher-level activities, thereby reducing the number of process steps and improving the overall readability of the model.

2. Enhancing model interpretability: By abstracting similar events into higher-level activities, event abstraction can help improve the interpretability of a process model by making it easier to understand the underlying business processes represented in the model.

3. Identifying process patterns: Event abstraction can help identify patterns in the process model that may not be immediately apparent when looking at individual events. This can help improve the model's manageability by identifying areas where process improvements may be needed.

4. Facilitating model comparison: Event abstraction can help facilitate the comparison of process models by providing a common language for representing similar process steps. This can help identify differences in process models derived from different event logs or for different business processes.

5. Supporting model refinement: Event abstraction can help support the refinement of a process model by providing a basis for identifying areas where additional process details may be needed. This can help improve the model's accuracy by ensuring that all relevant process steps are captured in the model.

In summary, by simplifying complex process models through event abstraction, it is possible to enhance their overall readability, interpretability, manageability, and accuracy, making them more effective tools for analyzing and optimizing business processes.