Event taxonomy
How the platform classifies what it ingests. The full category lists, with definitions and IDs, are in the reference pages linked below.
Why a standardized taxonomy exists
Different jurisdictions classify the same act differently. "Grand Larceny" in one department and "Theft from Person" in another describe the same underlying event.
The platform normalizes source-level labels into a standardized threat ontology using classification algorithms and, where needed, natural language processing. A category means the same thing everywhere, which is what makes a Chicago–Frankfurt comparison valid.
The structure
Events fall into two families:
| Family | What it covers | Reference |
|---|---|---|
| Crime | 26 categories across 3 groupings | Crime categories |
| Civil unrest | 25 event types across 3 groupings | Unrest categories |
Organizations using BaseFusion also have their own internal incident data classified on ingest. That mapping is defined per organization during setup — your account team will walk you through it.
Three crime groupings, not two
| Grouping | What it covers |
|---|---|
| Violent Crime | Force or the direct threat of force against a person |
| Property Crime | Taking or damaging property without direct violence |
| Regulatory Offenses | Offenses against law and regulation that are neither — fraud, corruption, weapons and drug violations, trafficking, organized crime |
Regulatory Offenses is the grouping people forget. Analysts who mentally split the world into violent and property crime miss it, and it carries categories that matter for financial services, compliance-driven organizations, and anyone assessing organized criminal activity.
Three unrest groupings
| Grouping | What it covers |
|---|---|
| Demonstrations | Protest activity, from peaceful gatherings to violent demonstrations |
| Political Violence | Organized violence involving armed actors — the highest-severity events in the taxonomy |
| Non-Violent Unrest Events | Strategic and organizational developments signalling changing conditions |
Severity weighting
Categories are not weighted equally. Violent crime is weighted well above property crime, which is why a location with a handful of serious violent incidents scores higher than one with a large volume of minor offenses. A homicide moves a score far more than a case of vandalism.
The exact per-category weighting is proprietary and is not published.
Unrest does not feed BaseScore
BaseScore is calculated from crime data only. Civil unrest events are tracked, categorized, and shown on the map, but they are not part of the score.
A location with significant unrest activity and low conventional crime can therefore carry a Low or Medium BaseScore. The score is not wrong — it is answering a narrower question than the map is. Where unrest matters to your assessment, read the unrest events directly.
See the BaseScore scale.
Subcategory scores
Not every category produces its own sub-score in the BaseScore breakdown. Drill-down happens only where data volume and coverage support it. Categories without a sub-score still contribute to the overall BaseScore — they just do not appear as a separate line.
If a category you care about is missing from a breakdown, this is usually why.
Filtering by category
Category filtering is available throughout the platform — on the map, in Trend Analysis, and in exports — and combines with date range and geographic filters.
This is how you answer a focused question: "how has vehicle theft trended at our distribution centers over the past year?"
Event categories are the one filter that changes the BaseScore, since the score is aggregated across the categories you have selected. A score captured with some categories switched off will not match the full score — worth noting when a figure needs to be reproducible. Other filters, including date range, do not move the score. See how scores update.
Overlapping event types
Some real-world events legitimately span categories. A protest that escalates into violence, or a demonstration where authorities detain participants, has elements of more than one.
The platform assigns a primary category based on the dominant characteristics of the event as reported. The unrest taxonomy handles much of this explicitly — Protest with intervention, Excessive force against protesters, and Violent demonstration are distinct categories precisely because the distinction matters.
If a categorization looks wrong for a specific incident, report it with the event and location.