Marisa Ferrara Boston

Marisa Ferrara Boston
research category image

Main area of research
Natural language & AI

Marisa Ferrara Boston is a Principal Research Scientist on the Question Answering team. Her research interest is in the area of Natural Language & AI. She joined Nuance after earning a Ph.D. in linguistics and cognitive science from Cornell University. At Nuance, Marisa has worked on various projects within the scope of question answering, including syntactic and semantic parsing, medical entity detection and annotation, and natural language inference. While at Cornell, she was involved in building computer models of human sentence processing, culminating in a dissertation that melds linguistic theory with psychological and computational approaches.

Selected articles

A computational model of cognitive constraints in syntactic locality

This dissertation is broadly concerned with the question: how do human cognitive limitations influence difficult sentences? The focus is a class of grammatical restrictions, locality constraints.

Read more

Parallel processing and sentence comprehension difficulty

Eye fixation durations during normal reading correlate with processing difficulty but the specific cognitive mechanisms reflected in these measures are not well understood. This study finds support

Read more

The role of memory in superiority violation gradience

This paper examines how grammatical and memory constraints explain gradience in superiority violation acceptability. A computational model encoding both categories of constraints is compared to experimental evidence. By

Read more

Dependency structures derived from minimalist grammars

This paper provides an interpretation of Minimalist Grammars [16,17] in terms of dependency structures. Under this interpretation, merge operations derive projective dependency structures, and movement

Read more

Parsing costs as predictors of reading difficulty: an evaluation using the Potsdam Sentence Corpus

The surprisal of a word on a probabilistic grammar constitutes a promising complexity metric for human sentence comprehension difficulty. Using two different grammar types, surprisal

Read more