Sciweavers

3371 search results - page 563 / 675
» Using parsimonious language models on web data
Sort
View
EMNLP
2009
15 years 2 months ago
The Feature Subspace Method for SMT System Combination
Recently system combination has been shown to be an effective way to improve translation quality over single machine translation systems. In this paper, we present a simple and ef...
Nan Duan, Mu Li, Tong Xiao, Ming Zhou
EMNLP
2007
15 years 6 months ago
Bootstrapping Feature-Rich Dependency Parsers with Entropic Priors
One may need to build a statistical parser for a new language, using only a very small labeled treebank together with raw text. We argue that bootstrapping a parser is most promis...
David A. Smith, Jason Eisner
CORR
2012
Springer
170views Education» more  CORR 2012»
14 years 6 days ago
What Cannot be Learned with Bethe Approximations
We address the problem of learning the parameters in graphical models when inference is intractable. A common strategy in this case is to replace the partition function with its B...
Uri Heinemann, Amir Globerson
ADCS
2004
15 years 5 months ago
On the Effectiveness of Relevance Profiling
Relevance profiling is a general process for withindocument retrieval. Given a query, a profile of retrieval status values is computed by sliding a fixed sized window across a doc...
David J. Harper, David Lee
EJC
2009
15 years 2 months ago
From Word Form Surfaces to Communication
The starting point of this paper is the external surface of a word form, for example the agent-external acoustic perturbations constituting a language sign in speech or the dots o...
Roland Hausser