Abstract

A construction and completion method of operational entity relationship knowledge graph (ERKG) is proposed to support tactical decision aids. The operational ERKG can be built by using a syntax indicator extraction method and an entity relationship extraction model combining syntax indicator with sentence-level context. And this ERKG can be complemented by employing a graph neural network inference model based on structured attention mechanism, and a relationship inference model combining operational entity knowledge with attention mechanism. While the entity relationship reasoning method combining the operational entity knowledge with attention mechanism is adopted to evaluate the trained relationship inference model. On the basis of structured ERKG, entity knowledge graph (EKG) and event knowledge graph (EventKG), tactical decision aid instructions can be generated by combining with the battlefield situation information.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call