Temporal knowledge graph
A temporal knowledge graph is a knowledge graph in which every fact carries time: when it was stated, when it became true, when it stopped being true. Knowledge is allowed to evolve — the graph keeps the history instead of overwriting it.
What is temporal knowledge graph?
Ordinary knowledge stores hold one version of the truth: the latest. That fails for real company knowledge, which changes and whose history is part of the knowledge. "We tune the line this way" matters less than "we tuned it this way after the 2018 failures, and here is what we did before and why we stopped." A temporal graph can answer how things are done now, how they were done then, and what changed in between.
Not to be confused with version history or backups. A backup can restore yesterday's data; it cannot answer a question about yesterday. In a temporal graph, time is queryable, not just recoverable.
Why it matters for SME continuity
Successors constantly relitigate old decisions because the reasoning behind them is gone — the company remembers the rule but not the incident that created it. With time in the graph, "why do we never ship partial orders to this client?" gets its 2014 answer, and the successor can judge whether the reason still holds instead of guessing.
Temporal knowledge graph inside MentX
The MentX knowledge base is a living, temporal knowledge graph. Every statement captured from a meeting or intake session carries its date, so DigiME can distinguish current practice from history — and can answer questions about the history itself, with sources.