Multilingual Dependency Parsing and Domain Adaptation using DeSR

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 feature selection to tune a feature model for each language. For the domain adaptation track we applied a tree revision method which learns how to correct the mistakes made by the base parser on the adaptation domain.

• Type: Conference
• Year: 2007
• Where: EMNLP
• Authors: Giuseppe Attardi, Felice dell'Orletta, Maria Simi, Atanas Chanev, Massimiliano Ciaramita
• Keywords: Domain Adaptation Track, EMNLP 2007, Incremental Deterministic Shift/reduce, Natural Language Processing, Order Averaged Perceptron