Anonymization

Anonymization is the process of removing, editing, obfuscating, and shuffling parts of the location data.

When data with location references is collected (raw data), the location data trajectories can reveal sensitive information about persons, devices, vehicles, etc.

The anonymization process is applied to remove the sensitive information from the collected data and to reduce the risk of unlawfully processing data and breaching the strict data privacy regulations employed around the world.

Anonymized data is different from the input data, for example:

Trajectories are split into multiple, separate, unordered sub-trajectories.
Unique identifiers are removed from trajectories.
Probe points are removed from the start, end, and/or middle of the trajectory or in certain areas.

Feature availability in operational modes

The following table shows the availability of anonymization methods in the two operational modes: streaming and batch.

To learn more, see Operational modes of HERE Anonymizer Self-Hosted.

Method	Streaming	Batch
Staypoint prediction	✅	✅
Smart gapping	✅	❌
Start-end cutting	✅	❌
Origin-destination obfuscation	❌	✅
Region selection exclusion	✅	✅
Region selection inclusion	✅	✅
Whitelisting	✅	✅
POI proximity data removal	✅	✅
Alerts detection and routing	✅	❌
Probe event handling	✅	✅
Density-aware anonymization	✅	❌

Precedence of features

You can enable multiple anonymization methods at the same time. In that case, the methods are applied to trajectories in the following order:

Whitelisting
POI proximity data removal
Region selection exclusion
Region selection inclusion
Density-aware anonymization
Origin-destination obfuscation / Start-end cutting
Smart gapping / Staypoint prediction

📘
Note
If whitelisting is applied, no other anonymization method is applied to the trajectory. To learn more, see Whitelisting.