Application of feature propagation to dependency parsing
Kepa Bengoetxea and Koldo Gojenola
11th International Conference on Parsing Technology (IWPT 2009)
Paris, France, 7th-9th October, 2009
Summary
This paper presents a set of experiments per-formed on parsing the Basque Dependency Treebank. We have applied feature propaga-tion to dependency parsing, experimenting the propagation of several of the most important morphosyntactic feature values from: a) auxil-iary verbs to the main verb, in verb phrases b) the last constituent to the head noun, in noun phrases c) the last conjunct to the conjunction, in coordination. In the experiments we have used the output of a parser to enrich the input of a second parser. Both parsers have been generated by Maltparser, a freely data-driven dependency parser generator. The transforma-tions, combined with the pseudoprojective graph transformation, obtain a LAS of 77.12% improving the best reported results for Basque.
START
Conference Manager (V2.56.8 - Rev. 780)