Wildfire Knowledge Graph · Evidence
A dated evidence state containing only information documented as available by that moment, without importing later findings.
In plain language
It answers what a defined observer could know then, rather than what the final investigation eventually concluded.
Why it matters
Separating event time from knowledge time prevents hindsight contamination in operational, public-information and forensic analysis.
Knowledge(t) = documented evidence published or available at or before tThe observer, time zone, source boundary and publication delay must be defined.
Known at a given time: the fundamental rule of an evidence timeline
‘Known at’ is an analytical question: what could a defined source, decision-maker or public audience reasonably have known by a particular timestamp? It prevents a reconstruction from using information published days or years later to judge an earlier decision as if the information had already existed.
Imagine a perimeter mapped at 10:00, uploaded publicly at 12:30 and analysed in a research report weeks later. The observation time, public availability and later analytical interpretation are three different moments. A claim that the map was ‘known’ at 10:00 requires evidence about who had access then, not only the timestamp of the underlying observation.
Suggested record fields: source organisation; observation or event time; publication or release time; archive/snapshot capture time; version identifier; audience and access limits; and the precise claim supported. If a time is missing, mark it ‘unknown’ rather than filling the gap with an unsupported value.
For public-facing incident analysis, use explicit phrases such as ‘recorded by the sensor at’, ‘published by the agency at’, or ‘reviewed by Tecnobosque on’. Those statements are more precise than a single unexplained ‘updated’ date.
Explore Evidence State, why records change and Wildfire Black Box.