Evann Smith is an experienced and innovative data scientist with substantial experience in research methodologies and algorithm development. Evann is an expert in natural language processing and machine learning, which Evann has leveraged to derive insights with applications to network science, economics and financial performance, political decision-making, and national security.
Dr. Smith has developed methodologies and applications using supervised and unsupervised methods, including a novel algorithm for engineering features from unstructured text that draws from both words embedding/neural network analyses and prototype driven learning in order to successfully derive features for classification in the most challenging of NLP data contexts—short and idiomatic texts, limited data, and rare vocabulary. Evann has developed agent-based models for predicting seemingly spontaneous mobilization across complex networks, created predictive data-driven typologies from latent variables, and analyzed over half a million emails to predict campaign success.
Dr. Smith holds a PhD in Government, with a focus in quantitative methodologies, from Harvard University. Prior to Evann's doctoral research, which analyzed the drivers of mass mobilization and the role of communication technologies in the Middle East, Evann received a Masters from the University of Chicago and a Bachelors of Arts with Honors from Columbia University. Throughout Evann's career in academia and industry, Dr. Smith has excelled in building innovative and data-driven solutions for real-world questions.