My primary research is in the area of natural language processing (NLP) where our goal is to develop algorithms and systems that will vastly improve a user's ability to find, absorb, and extract information from on-line text. My group's research generally proceeds at two complementary levels: we focus both on building real systems for large-scale natural language processing tasks and on developing techniques to address underlying theoretical problems in the syntactic, semantic and pragmatic analysis of natural language. As has become more or less standard in the field, we rely on statistical machine learning techniques including neural networks as our primary modeling tool, both for guiding natural language system development and for exploring the mechanisms that underlie language understanding. Our current work encompasses a number of related areas:

    argument mining, argument generation and argument summarization including identification of the logical structure of an argument and understanding the interaction between the persuasiveness of an argument and the beliefs of the reader;
    discourse-aware methods for opinion, event, and argument extraction;
    NLP methods for low-resource languages;
    neural representations for long documents.