Automated Input Processing for Machine Learning-Based Task Generation

Abstract

A method includes: detecting a plurality of events corresponding to operational disruptions for a facility, each event including one of a predetermined set of disruption type identifiers; retrieving configuration data defining, for each disruption type identifier, a set of impact attributes; determining, for each event, an impact score based on the set of impact attributes; generating, for each event, an exception record encoding the disruption type identifier and the impact score; sorting the exception records according to the impact scores; and providing the exception records, sorted according to the impact scores, to a task generator configured to output at least one task for responding to the operational disruptions.

Dependencies are inferred from the claim text.