Building a 'living database' for human-machine intelligence analysis

Abstract

In modern military operations the emphasis is on smaller teams and more ad hoc teamwork. This requires greater agility both in terms of capturing actionable intelligence as well as appropriate dissemination and fusion of that information to coalition team members based on their tasks and need to know. In previous research and development we have explored the potential for a controlled natural language (CNL), acting as the sole knowledge representation language to facilitate cooperative working among human and machine agents. In this paper we envision a 'living database' to support situation understanding, enable intelligence reporting, fusion and dissemination based on context. The human and machine users of the system have the ability to add, remove, or edit existing information using the CNL, including extensions to the model (or schema) in real-time. In this paper we describe various examples undertaken using this approach along with initial experiments using a conversational software agent to enable the field user to interact with such information in full natural English language.