Information Synthesis through Human-Machine Collaboration

Synthesis is the practice of integrating, organizing, filtering, evaluating, and compiling information such that it can be acted upon. Limitations on time, resources, and human memory place constraints on how much information can be synthesized by one person. This project explores how processes for synthesis may be distributed across people and augmented by machine learning techniques. We study how humans perform complex sense-making tasks like collecting, clustering, labelling, and re-representing unstructured information and we create workflows and tools to amplify this important and pervasive activity. Undergraduate or masters students who would like to aid in development of real-time interactive tools for collaborative synthesis and/or help us run experiments to test different approaches, please email your resume and a short description of your research experience to Steven Dow.