The key concept of the application is Timeline Set as a collection of timelines or timeline fragments defined by the set rules.
A set is associated with:
- Set definition rules – how the set is constructed.
- Metrics describing the set’s stats: duration, number of events, gaps in events, uniformity (evenness), etc. For each metrics like this we could have min, max, mean, median, 90 percentiles, standard deviation, etc.
- Metrics describing the set’s consistency in terms of topology: number (percent) of the unique sequences.
- Visualizations including such of the metrics, the timeline themselves (list, carousel) and the set as a whole (swimlane, flowchart)
Set Definition Rules
A set is defined by:
- Time window
- Fragmentation (all timeline fragments starting with N and ending on M)
- Milestones (events to show)
- Ignore events (events to hide)
- Any search criteria (timeline has N, them M, and no L in between)
- Collapse rules – ignore the repetition of a particular event
- Parallelism rules – ignore the order of some events which are performed in parallel
It is the full combination of rules which defines a set. User may add or remove any rule at any time.
A rule could be named and reused. For example, user could create a rule search rule “Event Surgery exists” and call this rule “Surgery patients”. Then the rule could be applied to multiple sets.
As users create filters and scope their analysis, sets can be saved to be viewed later.
Sets can be saved to a list to be retrieved and reviewed later.