介绍句法分析

[hailiang-wang]

介绍句法分析

依存关系主要使用句法分析器完成，句法分析器也有基于规则的和基于机器学习训练的。 虽然实现方案不同，但是二者的模型输出的是同样一个规范： https://nlp.stanford.edu/software/lex-parser.shtml

目前，state of art的可以参考 http://nlp.qq.com/semantic.cgi 句法类API

Google SyntaxNet

SyntaxNet 中文模型的使用

https://my.oschina.net/dingdayu/blog/1083438

How to train Chinese corpus after downloading the universal-dependencies-2.0 ?

https://github.com/dsindex/syntaxnet/issues/24

Google自然语言理解工具syntaxnet开放中文支持

https://www.jianshu.com/p/479a111ed5f4

HanLP

生成式依存句法分析器的简单实现

http://www.hankcs.com/nlp/to-achieve-a-simple-generative-dependency-parsing.html

基于神经网络的高性能依存句法分析器

http://www.hankcs.com/nlp/parsing/neural-network-based-dependency-parser.html

dependency parsing

论文分析 https://www.zhihu.com/question/46272554 https://arxiv.org/pdf/1603.06042v1.pdf

开源项目

https://gist.github.com/Samurais/2c935182fb7213205284f7cd5040536a https://github.com/Xe0n0/python-dependency-parser

Greedy Transition-Based Parsing

http://demo.clab.cs.cmu.edu/fa2015-11711/images/b/b1/TbparsingSmallCorrection.pdf https://www.youtube.com/results?search_query=Greedy+Transition-Based+Parsing

依存句法分析评价指标

https://jiaxuncai.github.io/2016/10/30/%E4%BE%9D%E5%AD%98%E5%8F%A5%E6%B3%95%E5%88%86%E6%9E%90%E8%AF%84%E4%BB%B7%E6%8C%87%E6%A0%87/

基于CRF序列标注的中文依存句法分析器的Java实现

http://www.hankcs.com/nlp/parsing/crf-sequence-annotation-chinese-dependency-parser-implementation-based-on-java.html

用于训练的数据集

universaldependencies

http://universaldependencies.org/

采用清华大学语义依存网络语料的20000句作为训练集。

http://www.hankcs.com/nlp/corpus/chinese-treebank.html#h3-6

• 汉语树库 http://www.hankcs.com/nlp/corpus/chinese-treebank.html

SyntaxNet Tutorial

https://github.com/tensorflow/models/blob/master/research/syntaxnet/g3doc/syntaxnet-tutorial.md

Conllu Data Format

https://github.com/taolei87/RBGParser/wiki/Data-Format

There is now a CoNLL-U data format as well which extends the CoNLL-X format.

CMU Neural Nets for NLP 2017 (12): Transition-based Dependency Parsing

http://www.phontron.com/class/nn4nlp2018/schedule/transition-parsing.html https://www.youtube.com/watch?v=7rp2c7JVymE

This lecture (by Graham Neubig) for CMU CS 11-747, Neural Networks for NLP (Fall 2017) covers:

• What is Transition-based Parsing?
• Shift-reduce Parsing w/ Feed-forward Nets
• Stack LSTM
• A Simple Alternative: Linearized Trees

Slides: http://www.phontron.com/class/nn4nlp2018/schedule/transition-parsing.html Code Examples: https://github.com/neubig/nn4nlp-code

Previous Video: https://youtu.be/e5sPNlgbZAE Next Video: https://youtu.be/gRtEW6Q5XJE

See more details of the class here: http://phontron.com/class/nn4nlp2017/

[hailiang-wang]

语法专业名词(部分)

https://wenku.baidu.com/view/b50dbedfe2bd960591c677d1.html