Deep Dive: Positive and Negative Times

Understanding the direction of time between process activities is essential for identifying anomalies, validating event sequences, and ensuring data integrity. The Positive and Negative Time Handling feature in Noreja enables advanced analysis by distinguishing whether transitions follow a natural forward timeline or reflect backward time jumps. This powerful filtering capability helps uncover data quality issues, detect reordered events, and isolate unusual process behavior—providing deeper insight into temporal dynamics within complex workflows.

In Noreja, we offer advanced filtering options that allow you to analyze process behavior based on the direction of time between activities – specifically whether the time between two activities is positive or negative.

This distinction helps you better understand unusual transitions, such as those where events happen out of chronological order or with suspiciously fast/slow timing.

What does "Time Handling" mean?

By default, our process graphs show all transitions between activities, regardless of whether the time between them is positive or negative. However, with time filtering enabled, you can focus specifically on transitions that follow a forward timeline (positive) or detect potential data/modeling issues that appear as backward time jumps (negative).

Positive Time Handling

If you enable Positive Time Handling, only transitions between activities are counted if their duration is greater than or equal to zero. This means that the second activity happened after the first activity, as expected.

Positive Time Handling

Example
In the transition from Mahnstufe 1 einleiten to Mahnstufe 2 einleiten, only those instances are included where the second activity occurred after the first.

Result:

  • 3 transitions from Mahnstufe 1 einleiten to Mahnstufe 2 einleiten
  • 1 transition from Mahnstufe 1 einleiten to Zahlung erhalten

Negative Time Handling

If you enable Negative Time Handling, the process graph shows only transitions where the duration is less than zero, meaning that the second activity happened before the first.

This mode is especially useful to:

  • Detect data quality issues (e.g. swapped timestamps)
  • Identify edge cases or event reordering
Negative Time Handling

Example
In the same case example, two Mahnstufe 1 einleiten activities show negative-time transitions to Mahnstufe 2 einleiten.
No transitions to Zahlung erhalten are counted in this mode.

Result:

  • 2 transitions from Mahnstufe 1 einleiten to Mahnstufe 2 einleiten
  • 0 transitions from Mahnstufe 1 einleiten to Zahlung erhalten

Why don’t the quantities add up?

You might wonder:
Why is the sum of positive and negative transitions not equal to the total number of transitions without any time filtering?

This happens due to what we call a Clique situation.
In some cases, one case instance might have multiple transitions to different successors, some with positive and others with negative durations. Since an activity can be counted in both directions, the same instance may appear in both positive and negative views – but only once in the unfiltered view.

In short:

  • The unfiltered view shows all transitions once.
  • The positive/negative filters may double-count some instances when looking at them separately.

Tip

You can toggle between positive and negative time filters in the Path Time Filter section on the right-hand panel within the Analyzer.
Use this to investigate unusual behavior, validate event orderings, or isolate time-based anomalies.

Was this article helpful?