My research focuses on making machine learning more useful, more interpretable, and able to learn and interact with humans. This helps users sift through decades of documents; discover when individuals lie, reframe, or change the topic in a conversation; or to compete against humans in games that are based on natural language.