Bias-Reducing Aptitudes and Needs Classifier for Honing ETL Recruitment (BRANCHER)

Army talent management processes are continually evolving to meet the needs of emerging challenges and technological advancements. A new initiative by the Army aims to identify uniformed experts, who will be designated as "Emerging Technology Leaders" (ETLs), to facilitate communication between research and operational communities. A solution that automates labor-intensive portions of this search effort would reduce the cost to identify, select, distribute, and develop the leaders that can immediately and continually excel at ETL positions for evolving and emerging communities. To meet these challenges, we propose to design and demonstrate the feasibility of a Bias-Reducing Aptitudes and Needs Classifier for Honing ETL Recruitment (BRANCHER). BRANCHER is an intelligent decision support system that identifies gaps and similarities between talent and job requirements, employing natural language processing to extract skills and a probabilistic knowledge graph to compare profiles, enabling assignment officers to fill current and future ETL designations while helping Soldiers plan career choices to suit these positions. BRANCHER ensures unbiased talent classification to provide a revolutionary approach to talent recruitment, selection, distribution, and development, combining a state-of-the-art neural network architecture for information extraction with a well-established theory of sociolinguistics, interactive career option recommendations, and continually-updated, crowdsourced data collection.