Jargon Association for Language Analysis in Domains (JALAD) II

For the DoD to maintain mission readiness and maximize organizational cohesion, leaders and decision makers at all levels must have an accurate view of personnel behaviors, perceptions, and sentiments. To support this goal, the DoD gathers extensive data from its personnel regarding organizational climate and opinions on work environment. For DoD analysts to use this large amount of text data effectively, they must be able to use and understand natural language processing (NLP) tools that will allow them to automate key parts of the data aggregation and analysis, thereby saving time and money. In a successful Phase I effort, Charles River Analytics designed Jargon Association for Language Analysis in Domains (JALAD), a text analysis tool that uses jargon term discovery and semantic association to improve NLP models for topic and sentiment analysis to easily draw understandable personnel insights from free-form text. JALAD provides the results of these NLP models through a Role Tailored Human-Machine Interface (HMI) that enables users without programing or NLP expertise to efficiently monitor, interpret, and leverage key summary features of free-form text data. For this Phase II effort, Charles River proposes to develop and evaluate a full-scope JALAD software prototype.