In this paper, we address the problem of extracting data records and their attributes from unstructured biomedical full text. There has been little effort reported on this in the ...
The construction of a speech recognition system requires a recorded set of phrases to compute the pertinent acoustic models. This set of phrases must be phonetically rich and balan...
We present a unified view of two state-of-theart non-projective dependency parsers, both approximate: the loopy belief propagation parser of Smith and Eisner (2008) and the relaxe...
This paper introduces algorithms for nonprojective parsing based on dual decomposition. We focus on parsing algorithms for nonprojective head automata, a generalization of head-au...
Terry Koo, Alexander M. Rush, Michael Collins, Tom...
In this paper we propose a new approach to improve electronic editions of human science corpus, providing an efficient estimation of manuscripts pages structure. In any handwriti...