|
Is it possible to develop such an intimate understanding of natural language that we can one day program a robot whose language abilities develop, in every circumstance, like a human infant's? In this
research we develop computational models of aspects of natural language and of the language-learning process. This research thus forms the intersection of linguistics and theoretical computer science. On the linguistics side, we draw from insights gained from theoretical syntax, semantics, morphology, phonology, and phonetics. On the computer science side, we draw from insights in formal language theory, formal learning theory, and grammatical inference. Current projects include: - Variation and learning stress patterns in the world's languages.
- Learning long distance agreement patterns (e.g., consonantal and
vowel harmony). - Learning phonotactic patterns simultaneously with speech
segmentation. - The sound pattern of Kwara'ae (Austronesian)
Phonology and phonetics lab
Principal investigator: Jeffrey Heinz |
