Colloque du
Département d'informatique et de recherche opérationnelle
Hal Daumé III
University of Maryland
Predicting Linguistic Structures Accurately and Efficiently
Résumé :
Many classic problems in natural language processing can be cast as
building mapping from a complex input (e.g., a sequence of words) to a
complex output (e.g., a syntax tree or semantic graph). This task is
challenging both because language is ambiguous (learning difficulties)
and represented with discrete combinatorial structures (computational
difficulties). Often these are at odds: the features you want to add
to decrease learning difficulties cause nontrivial additional
structure yielding worse computational difficulties.
I will begin by discussing algorithms that side-step the issue of
combinatorial blowup and aim to predict an output structure directly.
I will then present approaches that explicitly learn to trade-off
accuracy and efficiency, applied to a variety of linguistic phenomena.
Moreover, I will show that in some cases, we can actually obtain a
model that is faster and more accurate by exploiting smarter learning
algorithms.
This is joint work with Jason Eisner, John Langford,
Jiarong Jiang, He He, Adam Teichert and Tim Vieira.
Le jeudi 21 novembre 2013, 15h30, pavillon André-Aisenstadt, salle 3195
Information : Gilles Brassard, brassard@iro.umontreal.ca