Information Extraction Using Controlled English to Support Knowledge-Sharing and Decision-Making

Abstract : Current and future coalition operations involve multi-team and/or multi-nation collaborations. While large volumes of structured/unstructured data are often available, improvement of data access, information extraction, and knowledge sharing is critically important but remains a major challenge for effective and efficient C2 operations. In this paper, we propose an approach to information extraction using International Technology Alliance Controlled English (CE) to improve fact extraction and knowledge sharing, aiming to enhance situation awareness and support decision-making. CE is a subset of English with a restricted grammar to reduce complexity and avoid ambiguity. The current version of CE has a formal syntax and semantics and is consistent with First Order Predicate Logic. CE is used to model both the inputs and outputs of the information extraction process, and to support end-users in configuring information extraction tools. Thus, CE provides, among other things (i) A user-friendly language for queries and system-to-user report representation. (ii) A common form of expression that supports extending and modifying domain models (ontologies), and enables mapping between models and terminology or language variants. CE-based information extraction will greatly facilitate the processes in the cognitive and social domains that enable forces with diverse backgrounds to collaborate effectively and efficiently.