Probabilistic Models for High-Order Projective Dependency Parsing
This work addresses dependency parsing for multiple languages, offering incremental improvements in computational efficiency and performance.
The paper tackles high-order projective dependency parsing by developing generalized probabilistic models and efficient algorithms for computing partition functions and marginals, extending the inside-outside algorithm. Experiments on English, Chinese, and Czech show competitive performance for English and state-of-the-art results for Chinese and Czech.
This paper presents generalized probabilistic models for high-order projective dependency parsing and an algorithmic framework for learning these statistical models involving dependency trees. Partition functions and marginals for high-order dependency trees can be computed efficiently, by adapting our algorithms which extend the inside-outside algorithm to higher-order cases. To show the effectiveness of our algorithms, we perform experiments on three languages---English, Chinese and Czech, using maximum conditional likelihood estimation for model training and L-BFGS for parameter estimation. Our methods achieve competitive performance for English, and outperform all previously reported dependency parsers for Chinese and Czech.