Description: SDi has designed an architecture, in accordance with the parameters of N202_126, which utilizes synthetic text generation to simplify the process of scenario inject creation by providing the exercise controller with a software tool that both creates and stores scenario injects. This architecture utilizes Natural Language Processing and Pattern-Exploiting Training to develop user-requested text which maintains the specific properties (grammar, subject, and sentiment) of human-written text. Social Synthesis’ process of Intelligent Context Matching (ICM) is a new hybrid technique developed from proven NLP processes and is designed to be both computationally “lightweight” and simple to implement.
Primary Objective: SDi is developing Social Synthesis to provide Mission Scenario Event List (MSEL) items for use in military and crisis scenario exercises.
Innovation: Social Synthesis utilizes the SDi-developed Intelligent Context Matching system, which allows the generation of synthetic text that can match simple human-generated “starter text” to create a more complex narrative.