Paper: Automatic Treebank Conversion via Informed Decoding

ACL ID C10-2176
Title Automatic Treebank Conversion via Informed Decoding
Venue International Conference on Computational Linguistics
Session Poster Session
Year 2010

In this paper, we focus on the challenge of automatically converting a constituency treebank (source treebank) to fit the stan- dard of another constituency treebank (tar- get treebank). We formalize the conver- sion problem as an informed decoding procedure: information from original an- notations in a source treebank is incorpo- rated into the decoding phase of a parser trained on a target treebank during the parser assigning parse trees to sentences in the source treebank. Experiments on two Chinese treebanks show significant im- provements in conversion accuracy over baseline systems, especially when training data used for building the parser is small in size.