We present several improvements to unlexicalized parsing with hierarchically state-split
PCFGs. First, we present a novel coarse-to-fine method in which a grammar's own …

We show how features can easily be added to standard generative models for unsupervised
learning, without requiring complex new training methods. In particular, each component …

ﾇ ﾈ ktyd ﾉ sEFHxC ﾊ nm) ﾋﾌ ﾍ h ﾎ慫 ﾏ I ﾐ厶ｦﾑｦﾒ W ﾓ ﾔ f ﾕ x ﾒ w ﾍ h ﾎび 4 ﾐ曼 h ﾎ ﾕ x ﾒ x
ﾏ沌対 x ﾘ v ﾗｦﾔ r ﾑｦﾎば ﾏ I ﾐ燮ｦﾚ [ﾑ 4 ﾐ h ﾛ q ﾛﾜﾐ h ﾏ v ﾔﾖ x ﾚ I ﾐ曼 h ﾎｿ ﾖ h ﾑ 4 ﾐ h …

Various machine learning tasks can benefit from access to external information of different
modalities, such as text and images. Recent work has focused on learning architectures with …

While rules and exemplars are usually viewed as opposites, this paper argues that they form
end points of the same distribution. By representing both rules and exemplars as (partial) …

There is precisely one complete language processing system to date: the human brain.
Though there is debate on how much built-in bias human learners might have, we definitely …

We introduce two first-order graph-based dependency parsers achieving a new state of the
art. The first is a consensus parser built from an ensemble of independently trained greedy …