Publication at Faculty of Mathematics and Physics |
2015
Abstract
The paper describes a series of experiments on building a dependency parsing model using MaltParser, the SynTagRus treebank of Russian, and the morphological tagger Mystem. The experiments have two purposes.
The first one is to train a model with a reasonable balance of quality and parsing time. The second one is to produce user-friendly software which would be practical for obtaining quick results without any technical knowledge (programming languages, linguistic tools, etc.).