Atanas Chanev

Learn More
One of the potential advantages of data-driven approaches to natural language processing is that they can be ported to new languages, provided that the necessary linguistic data resources are available. In practice, this advantage can be hard to realize if models are overfitted to a particular language or linguistic annotation scheme. Thus, using two(More)
We describe our experiments using the DeSR parser in the multilingual and domain adaptation tracks of the CoNLL 2007 shared task. DeSR implements an incremental deterministic Shift/Reduce parsing algorithm, using specific rules to handle non-projective dependencies. For the multilingual track we adopted a second order averaged perceptron and performed(More)
Multilingual dependency parsing is gaining popularity in recent years for several reasons. Dependency structures are more adequate for languages with freer word order than the traditional constituency notion. There is a growing availability of dependency treebanks for new languages. Broad coverage statistical dependency parsers are available and easily(More)
Natural Language Processing (NLP) is one of the most challenging fields in AI. Machine translators and other practical tools have been implemented recently for wide spread languages like English, German etc. The natural languages are quite different from each other and that makes the usage of models for English in systems, processing less spoken languages a(More)
We present a translation model based on dependency trees. The model adopts a treeto-string approach and extends PhraseBased translation (PBT) by using the dependency tree of the source sentence for selecting translation options and for reordering them. Decoding is done by translating each node in the tree and combining its translations with those of its(More)
  • 1