Knowledge-directed AI Reasoning Over Schemas (KAIROS)
Abstract
The Knowledge-directed Artificial Intelligence (AI) Reasoning over Schemas (KAIROS) program will develop AI and machine learning technologies to aid a human operator in understanding complex event sequences. For the purposes of KAIROS, an event is an occurrence that results in an observable and recognizable change in either the physical world or human society. Events of particular interest to KAIROS are those that create changes that have significant impact on national or homeland security. Many important events are not simple occurrences but complex phenomena that are composed of numerous subsidiary event elements, some of which happen simultaneously while others are sequential and dependent on each other. Humans make sense of event sequences by organizing them into narrative structures that may occur or re-occur frequently. These structures are abstracted into schemas - organized units of knowledge that represent patterns - for the purpose of cognition. The KAIROS program will develop automated systems that use existing schemas and, when needed, create new schemas to bring structure to complex event sequences and present these structured representations to operators. Given multi-media inputs, operators will use KAIROS technologies to identify subsidiary event elements, determine their temporal order, recognize complex event sequences, and link disparate events. KAIROS technologies will enable analysts and warfighters to understand unfolding events rapidly and accurately.
Document Details
- Document Type
- Accomplishment
- Publication Date
- Oct 01, 2020
- Source ID
- ed3f23d3f234d6b66acf17161c69d099