Reference Number: ARL-SRP-00067

Project Description

This project seeks to discover how humans trust (or distrust) what they read in today’s complex information environment. This project will conduct research into how to present evidence of deception found in natural language texts. A critical element of this work will be how to provide explanations of evidence to gain, not sacrifice, the human’s trust. The internship will involve the scoping of a research question that fits within these areas of deception, trust, and user interaction paradigms for labeling or exploration. The intern will design an actionable research plan, conduct the plan and adapt to unforeseen circumstances, and document findings in an academic paper draft. Possible focused topics include developing labeled corpora, experimenting with signature models or developing novel workflows, implementing proof-of-concept dashboards and interaction paradigms, and/or designing evaluation criteria for assessing state-of-the-art and new models. 

Technical Skills

• Computer Science
• Artificial Intelligence/Machine Learning
• Natural Language Processing            
• Interest and/or experience in computer science, human-agent interaction, and natural language processing