My primary research interests lie in natural language understanding, knowledge discovery and representation, computational recognition of salient information in text, and uncertainty management. These interests span Artificial Intelligence, particularly Natural Language Processing and Cognitive Science, as well as Imprecision Management.
Recent advances in AI have transformed how people interact with computational systems, making informal, natural-language communication increasingly routine. Yet the apparent ease of these interactions can obscure challenges that have long been central to my research: navigating between different levels of abstraction, communicating at the appropriate level, understanding what is stated as well as what is left implicit, accommodating fuzziness, vagueness, and uncertainty where appropriate, and maintaining factual precision where required. As interactions with computational systems become more continuous and informal, privacy becomes part of this problem as well: natural language contains substantial information about its users, and information that may appear innocuous in isolation can reveal sensitive information when accumulated and considered together. My research seeks to understand how computational models can address these interconnected challenges in ways that align with human communication, expectations, and needs. Computational humor provides both a particularly visible entry point into these questions and a demanding testbed for exploring them.
