下载  >  人工智能  >  机器学习  > eisenstein nlp notes

eisenstein nlp notes 评分

分享 乔治亚理工大学 Jacob Eisenstein 教授开放了自然语言处理领域的最新教材《Natural Language Processing》。 这本书的内容主要分为四大章节,即 NLP 中监督与无监等学习问题、序列与解析树等自然语言的建模方式、语篇语义的理解,以及后这些技术最在信息抽取、机器翻译和文本生成等具体任务中的应用。 Eisenstein 将这一本非常完善的教材称之为「Notes」,它是在乔治亚理工大学学习自然语言处理相关领域所需要了解的基础。例如在介绍自然语言处理理论与方法的课程 CS4650/7650 中,这本开放书籍就作为标准的课程教材。
Contents Contents 61 Introduction 13 1.1 Natural language processing and its neighbors ,.13 1. 2 Three themes in natural language processing 17 1. 2.1 Learning and knowledge 17 1.2.2 Search and learning 1.2.3 Relational, compositional, and distributional perspectives 1. Learning to do natural language processing.. hal p 20 22 1.32 How to usc this book...…··.. 1.3.1 Background 14 15 I Learning 27 16 2 Linear text classification 29 2.1 Naive b 32 2.1.1 Types and tokens 34 2.1.2 Prediction 35 2.1.3 Estimation 36 21 2.1.4 Smoothing and map estimation 2.1.5 Setting hyperparameters 38 2.2 Discriminative learning ,,.39 2.2.1 Perceptron 40 2.2.2 Averaged perceptron 42 2.3 Loss functions and large-margin classification 2.3.1 Large margin classification 46 2.3.2 Support vector machines 47 2.3.3 Slack variables 2.4 Logistic regression 50 2.4.1 Regularization 51 CONTENTS 2.4.2 Gradients 52 2.5 Optimization 52 2.5.1 Batch optimization 53 2.5.2 Online optimization 54 2.6 *Additional topics in classification 56 2.6.1 Feature selection by regularization 2.6.2 Other views of logistic regression 56 2.7 Summary of learning algorithms 58 40 3 Nonlinear classification 61 4 3.1 Feedforward neural networks 2 3. 2 Designing neural networks 64 3. 2.1 Activation functions 44 3.2.2 Network structure 3.2.3 Outputs and loss functions 3. 2. 4 Inputs and lookup layers 67 3.3 Learning neural networks · 67 3.3.1 Backpropagation 3.3.2 Regularization and dropout 3.33 earning theory····: 3.3.4 Tricks 3.4 Convolutional neural networks 53 4 Linguistic applications of classification 81 4.1 Sentiment and opinion analysis ,,81 4.1.1 Related problems 83 4.1.2 Alternative approaches to sentiment analysis 84 4.2 Word sense disambiguation 4.2.1 How many word senses? 4.2.2 Word sense disambiguation as classification 87 4.3 Design decisions for text classification 88 4.3.1 What is a word? .88 4.3.2 How many words? 91 4.3.3 Count or binary? 92 4.4 Evaluating classifiers 92 65 4.4.1 Precision recall, and f-measure 93 4.4.2 Threshold-free metrics 95 4.4.3 Classit assifier comparison and s tatistical significance 4.4.4 Multiple comparisons 4.5 Building datasets Qc 4.5.1 Metadata as labels ,,,,,,,,100 (c) Jacob Eisenstein 2018. Draft of June 1, 2018 CONTENTS 71 4.5.2 Labeling data 10 72 5 Learning without supervision 107 5.1 Unsupervised learning 107 5.1.1 K-means clusterin 108 5.1.2 Expectation Maximization(Em) 110 5.1.3 EM as an optimization algorithm 114 5.1.4 How many clusters? ..115 5.2 Applications of expectation-maximization ..116 5.2.1 Word sense induction 5.2.2 Semi-supervised learning 117 5.2.3 Multi-component modeling 118 5. 3 Semi-supervised learning ····.·······鲁 5.3.1 Multi-view learning 120 5.3.2 Graph-based algorithms 121 5.4 Domain adaptation ,,.,122 5.4.1 Supervised domain adaptation 123 5.4.2 Unsupervised domain adaptation 124 5.5 *Other approaches to learning with latent variables 126 890 5.5.1Sam ling pi 126 5.5.2 Spectral learning .128 91 II Sequences and trees 135 92 6 Language models 137 6.1 N-gram language models .··· 138 6.2 Smoothing and discounting 141 95 6.2.1 Smoothing 141 6.2.2 Discounting and backoff 142 6.2.3 *Interpolation ,,,,,,,,,,,,,,,,,143 6.2. 4 Kneser-Ney smoothing 6.3 Recurrent neural network language models 146 100 6.3.1 Backpropagation through time 148 6.3.2 Hyperparameters 149 102 6.3. 3 Gated recurrent neural networks 149 6.4 Evaluating language models ·:.·· ..151 6.4.1 Held-out likelihood 151 6.4.2 Perplexity 152 106 6.5 Out-of-vocabulary words 153 (c Jacob Eisenstein 2018. Draft of June 1, 2018 CONTENTS 107 7 Sequence labeling 155 7.1 Sequence labeling as classification 155 7.2 Sequence labeling as structure prediction 157 110 7.3 The Viterbi algorithm 159 111 7.3.1 Example 162 112 7.3.2 Higher-order features 163 113 7. 4 Hidden markov Models 163 114 7.4.1 Estimation 165 115 7.4.2 Inference 7.5 Discriminative sequence labeling with features .167 117 7.5.1 Structured perceptron 170 118 7.5.2 Structured support vector machines ..170 119 7.5.3 Conditional random fields 7.6 Ncural scqucncc labeling 177 121 7.6.1 Recurrent neural networks ··, 177 7.6.2 Character-level models 7.6.3 Convolutional Neural Networks for Sequence labeline .179 180 7.7 * Unsupervised sequence labeling .180 7.7.1 Linear dynamical systems .182 7.7.2 Alternative unsupervised learning methods 182 127 7.7.3 Semiring Notation and the Generalized Viterbi Algorithm ..... 182 8 Applications of sequence labeling 185 8.1 Part-of-speech tagging 130 8.1.1 Parts-of-Speech 186 8.1.2 Accurate part-of-speech tagging 190 132 8.2 Morphosyntactic Attributes 192 8.3 Named Entity Recognition 193 134 8.4 Tok 195 135 8.5 Code switching 196 136 8.6 Dialogue acts 1379 Formal language theory 199 138 9.1 Regular languages 200 9.1.1 Finite state 201 140 9.1.2 Morphology as a regular language 202 141 9.1.3 Weighted finite state acceptors 204 142 9. 1.4 Finite state transducers 209 143 9.1.5 *Learning weighted finite state automata 214 144 9.2 Context-free languages 215 145 9.2.1 Context-free grammars .,,,,,,,,,,216 (c) Jacob Eisenstein 2018. Draft of June 1, 2018 CONTENTS 146 9. 2.2 Natural language syntax as a context-free language .219 147 9.2.3 A phrase-structure grammar for English ..221 148 9.2.4 Grammatical ambiguity 226 9.3 *Mildly context-sensitive languages 226 150 9.3.1 Context-sensitive phenomena in natural language........... 227 151 9.3.2 Combinatory categorial grammar 228 152 10 Context-free parsing 233 10.1 Deterministic bottom-up parsing ,,,234 154 10.1.1 Recovering the parsc trcc 236 155 10.1.2 Non-binary productions .236 156 10.1.3 Complexity 237 157 10.2 Ambiguity 237 158 10.2.1 Parser evaluation 238 10.2.2 Local solutions ,,,,,,,,,239 10.3 Weighted Context-Free grammars 240 10.3.1 Parsing with weighted context-free grammars .241 10.3.2 Probabilistic context-free grammars 243 10.3.3 *Semiring weighted context-free grammars 245 164 10.4 Learning weighted context-free grammars ,,.245 10.4. 1 Probabilistic context-free grammars 246 166 10.4.2 Feature-based parsing 246 10.4.3 *Conditional random field parsing............. 247 10.4.4 Neural context-free grammars ..................... 249 169 10.5 Grammar refinement ,,.250 10.5. 1 Parent annotations and other tree transformations 251 171 10.5.2 Lexicalized context-free grammars 252 172 10.5.3 *Refinement grammars 256 173 10.6 Beyond context-free parsing 257 174 10.6. 1 Reranking 257 175 10.6.2 Transition-based parsing ,,,,258 176 11 Dependency parsing 261 177 11.1 Dependency grammar 261 178 11.1.1 Heads and dependents 262 179 11.1.2 Labeled dependencies ,..263 180 11.1.3 Dependency subtrees and constituents 264 11.2 Graph-based dependency parsing 266 18 11.2.1 Graph-based parsing algorithms 268 11.2.2 Computing scores for dependency arcs ,269 11.23 Learning,,,,,,,,,,,, 271 (c Jacob Eisenstein 2018. Draft of June 1, 2018 CONTENTS 11.3 Transition-based dependency parsing 272 11.3.1 Transition systems for dependency parsing .273 187 11.3.2 Scoring functions for transition-based parsers 277 11.3.3 Learning to parse ,278 11. 4 Applications 281 190 III Meaning 285 191 12 Logical semantics 287 192 12.1 Meaning and denotation .288 12.2 Logical representations of meaning ..289 194 12.2.1 Propositional logic 289 195 12.2.2 First-order logi IC 290 12.3 Semantic parsing and the lambda calculus ,.294 197 12.3.1 The lambda calculus 295 198 12.3.2 Quantification 297 12. 4 Learning scmantic parsers ......... ,,299 200 12.4.1 Learning from derivations 300 12.4.2 Learning from logical forms 302 12.4.3 Learning from denotations ..303 203 13 Predicate-argument semantics 309 204 13.1 Semantic roles 311 13.1.1 VerbNet 312 13. 1.2 Proto-roles and Prop Bank 313 207 13.1.3 FrameNet 314 208 13.2 Semantic role labeling ..316 209 13.2.1 Semantic role labeling as classification .316 210 13.2.2 Semantic role labeling as constrained optimization 319 211 13.2.3 Neural scmantic rolc labeling ..321 13.3 Abstract Meaning representation 322 213 13.3.1 AMR Parsing ..325 214 13.4 Applications of Predicate-Argument Semantics 326 215 14 Distributional and distributed semantics 333 216 14.1 The distributional hypothesis 333 217 14.2 Design decisions for word representations 335 14.2.1 Representation ..335 219 14.2.2 Context 336 220 14.2.3 Estimation 337 (c) Jacob Eisenstein 2018. Draft of June 1, 2018 CONTENTS 14.3 Latent semantic analysis .338 222 14.4 Brown clusters ,,,,,,339 223 14.5 Neural word embeddings 343 14.5.1 Continuous bag-of-words(CBOw) 343 14.5.2 Skipgrams 344 14.5.3 Computational complex ,,344 14.5.4 Word embeddings as matrix factorization 346 28 14.6 Evaluating word embeddings 347 229 14.6.1 Intrinsic evaluations 347 230 14.6.2 Extrinsic evaluations 348 31 14.7 Distributed representations beyond distributional statistics 349 14.7. 1 Word-internal structure ..350 14.7.2 Lexical semantic resources 352 14.8 Distributed representations of multiword units 352 235 14.8.1 Purely distributional methods 352 36 14.8.2 Distributional-compositional hybrids 353 237 14.8.3 Supervised compositional methods .354 238 14.8. 4 Hybrid distributed-symbolic representations 355 239 15 Reference resolution 359 240 15. 1 Forms of referring expressions 360 241 15.1.1 Pronouns 360 242 15.1.2 Proper nouns 365 15.1.3 Nominals ·:···· 366 244 15.2 Algorithms for coreference resolution 366 15.2.1 Mention-pair models 367 15.2.2 Mention-ranking models ..368 15.2.3 Transitive closure in mention-based models 369 248 15.2.4 Entity-based models ..370 249 15.3 Representations for coreference resolution 375 250 15.3.1 Features,.,,,,, 376 15.3.2 Distributed representations of mentions and entities.......... 378 252 15.4 Evaluating coreference resolution 381 253 16 Discourse 385 254 16.1 Segments 385 255 16.1.1 Topic segmentation 386 256 16.1.2 Functional segmentation 387 257 16.2 Entities and reference 387 16.2. 1 Centering theory 388 259 16.2.2 The entity grid 389 (c Jacob Eisenstein 2018. Draft of June 1, 2018 CONTENT 260 16.2.3 *Formal semantics beyond the sentence level ..390 16. 3 Relations .390 16.3.1 Shallow discourse relations 391 263 16.3.2 Hierarchical discourse relations 394 16.3.3 Argumentation ...398 16.3.4 Applications of discourse relations 399 266 IV Applications 405 267 17 Information extraction 407 17.1 Entities 409 17.1.1 Entity linking by learning to rank 410 270 17.1.2 Collective entity linking .412 17. 1.3 *Pairwise ranking loss functions 413 272 17.2 Relations 415 273 17.2.1 Pattern-based relation extraction .416 17.2.2 Relation extraction as a classification task .............. 417 275 17.2.3 Know ledge base population 420 276 17.2.4 Open information extraction 424 17.3 Events .425 17. 4 Hedges, denials, and hy potheticals 426 279 17.5 Qucstion answering and machinc reading ..428 280 17.5.1 Formal semantics 428 17.5.2 Machine reading .429 282 18 Machine translation 435 1 8.1 Machine translation as a task 435 284 18.1.1 Evaluating translations .437 18.1.2Data .439 18.2 Statistical machine translation 440 18.2.1 Statistical translation modelin 441 18.2.2 Estimation ..443 18.2. 3 Phrase-based translation 444 290 18.2.4 *Syntax-based translation 445 18.3 Neural machine translation 446 18.3.1 Neural attention 448 18.3.2 *Neural machine translation without recurrence 450 294 18.3.3 Out-of-vocabulary words ..451 18.4 Decoding 453 296 18.5 Training towards the evaluation metric 45 (c) Jacob Eisenstein 2018. Draft of June 1, 2018

...展开详情
所需积分/C币:2 上传时间:2018-07-22 资源大小:14.3MB
举报 举报 收藏 收藏
分享 分享
Natural Language Processing by Jacob Eisenstein

nlp学习的基础教材,共583页,内含目录。

立即下载
Natural Language Processing Jacob Eisenstein

乔治亚理工大学的 Jacob Eisenstein 教授开放了自然语言处理领域的最新教材《Natural Language Processing》,该教材于 2018 年 6 月开放第一版。

立即下载
Jacob Eisenstein -natural language processing notes

乔治亚理工大学 Jacob Eisenstein 教授开放了自然语言处理领域的最新教材《Natural Language Processing》,该教材在 2018 年 6 月完成

立即下载
nlp-intent-toolkit, 基于OpenNLP的意向识别.zip

nlp-intent-toolkit, 基于OpenNLP的意向识别 nlp-intent-toolkit使用OpenNLP的插槽识别意图。这是使用OpenNLP训练系统接受自然语言输入,特别是通过speech-to-text源进行输入,并返回带有参数的识别操作的一个例子。 系统使用文档分类确定输入和实

立即下载
cs229-notes10.pdf

cs229-notes10.pdf

立即下载
GA-NLP-MATLAB

基于遗传算法和非线性规划的函数最优化求解

立即下载
理工四大

学校怀旧

立即下载
超级影响力-NLP致胜行销学.pdf

跨部门沟通的重要性 问题一:作为部门主管,是沟通能力重要还是其它能力重要? 良好的人际关系者,可使工作成功与个人幸福 获得率达85%以上 针对10000人的记录进行分析,成功的因素中 85%决定于人际关系,而知识、技术、经验只 占15% 某年度某地区被解雇的4000人中,不称职的 占10%,人际关系不好者则占90% 根据5年的跟踪调查,人际关系好的人平均年 薪比优等生高15%,比劣等生高33% 合格的部门主管,沟通能力应占80%, 而其它能力只需20%。 沟通能力强的人更善于管理。 目前大多数企业的中,高层主管最缺乏的是沟通能力。 沟通能力强的正确观念与心态。 某企业总助被解职的案例。

立即下载
ModbusTCP/RTU网关设计

基于UIP协议栈,实现MODBUS联网,可参考本文档资料,有MODBUS协议介绍

立即下载
html+css+js制作的一个动态的新年贺卡

该代码是http://blog.csdn.net/qq_29656961/article/details/78155792博客里面的代码,代码里面有要用到的图片资源和音乐资源。

立即下载
iCopy解码软件v1.0.1.7.exe

解ic,id,hid卡密码破解ic,id,hid卡密码破解ic,id,hid破解ic,id,hid卡破解ic,id,hid卡密码密码卡密码破解ic,id,hid卡...

立即下载
分布式服务框架原理与实践(高清完整版)

第1章应用架构演进1 1.1传统垂直应用架构2 1.1.1垂直应用架构介绍2 1.1.2垂直应用架构面临的挑战4 1.2RPC架构6 1.2.1RPC框架原理6 1.2.2最简单的RPC框架实现8 1.2.3业界主流RPC框架14 1.2.4RPC框架面临的挑战17 1.3SOA服务化架构18 1.3.1面向服务设计的原则18 1.3.2服务治理19 1.4微服务架构21 1.4.1什么是微服务21 1.4.2微服务架构对比SOA22 1.5总结23 第2章分布式服务框架入门25 2.1分布式服务框架诞生背景26 2.1.1应用从集中式走向分布式.26?

立即下载
Camtasia 9安装及破解方法绝对有效

附件中注册方法亲测有效,加以整理与大家共享。 由于附件大于60m传不上去,另附Camtasia 9百度云下载地址。免费自取 链接:http://pan.baidu.com/s/1kVABnhH 密码:xees

立即下载
电磁场与电磁波第四版谢处方 PDF

电磁场与电磁波第四版谢处方 (清晰版),做天线设计的可以作为参考。

立即下载
压缩包爆破解密工具(7z、rar、zip)

压缩包内包含三个工具,分别可以用来爆破解密7z压缩包、rar压缩包和zip压缩包。

立即下载
source insight 4.0.0087 注册机序列号Patched(2017/10/17)

最新的sourceinsight4.0.0087和谐license及和谐文件。真正的4087版本,使用附件中的license文件,替换sourceinsight4.exe

立即下载
Java项目经验汇总(简历项目素材)

Java项目经验汇总(简历项目素材)

立即下载
支付宝转账demo-实现提现功能

支付宝单笔转账,实现提现功能,自己用到,所以分享给大家,希望可以帮到大家。(内有demo实例,望大家多多提意见)

立即下载
算法第四版 高清完整中文版PDF

《算法 第4版 》是Sedgewick之巨著 与高德纳TAOCP一脉相承 是算法领域经典的参考书 涵盖所有程序员必须掌握的50种算法 全面介绍了关于算法和数据结构的必备知识 并特别针对排序 搜索 图处理和字符串处理进行了论述 第4版具体给出了每位程序员应知应会的50个算法 提供了实际代码 而且这些Java代码实现采用了模块化的编程风格 读者可以方便地加以改造

立即下载