This is a brief summary of paper for me to study and organize it, A Fast and Accurate Dependency Parser using Neural Networks (Chen and Manning., EMNLP 2014) I read and studied.

This paper propose the model based on neural network for dependecy parsing task in NLP.

They address the problem that feature is sparsness and depends on expertise for developer for depedency parsing task or something in NLP.

Especially back then using property that low-dimensional, dense word embedding can effectively alleviate sparsity by sharing statistical strength between similar wors, and can provide them a good starting poit to construct features of words and their interactions.

Chen and Manning., EMNLP 2014

They propose the denpendecy parsing system with transition-based arc-standard, briefly saying how for the arc-standard-based transition-based dependency parsing to work.

Chen and Manning. EMNLP 2014

Also they showed a variety of model analysis: Pos tag and arc label embedding can capture semantic information very well in Section 4.4 expecting the possiblity to use it in other NLP task.

Note(Abstract): Almost all current dependency parsers classify based on millions of sparse indicator features. Not only do these features generalize poorly, but the cost of feature computation restricts parsing speed significantly. In this work, they propose a novel way of learning a neural network classifier for use in a greedy, transition-based dependency parser. Because this classifier learns and uses just a small number of dense features, it can work very fast, while achieving an about 2% improvement in unlabeled and labeled attachment scores on both English and Chinese datasets. Concretely, our parser is able to parse more than 1000 sentences per second at 92.2% unlabeled attachment score on the English Penn Treebank.

Download URL:
The paper: A Fast and Accurate Dependency Parser using Neural Networks (Chen and Manning., EMNLP 2014)

Reference

Paper
- EMNLP Version: A Fast and Accurate Dependency Parser using Neural Networks (Chen and Manning., EMNLP 2014)
How to use html for alert
- how to use icon

A Fast and Accurate Dependency Parser using Neural Networks

Title of paper - A Fast and Accurate Dependency Parser using Neural Networks

A Fast and Accurate Dependency Parser using Neural Networks

Title of paper - A Fast and Accurate Dependency Parser using Neural Networks

Reference