Probabilistic Allen algebra lets you reason about uncertain temporal relations from language and data by deriving relation probabilities from distributions over interval boundaries, rather than assigning scores arbitrarily—making temporal reasoning more grounded in uncertainty.
This paper extends Allen's interval algebra—a system for reasoning about temporal relationships—to handle uncertainty. Instead of crisp yes/no answers about whether one event is "before" another, it models time points and interval boundaries as probability distributions (Gaussians), allowing expressions like "roughly during" or "just before" to have graded meanings.