| 自然语言处理(NLP)是人工智能研究中极具挑战的一个分支。随着深度学习等技术的引入,NLP 领域正在以前所未有的速度向前发展。但对于初学者来说,这一领域目前有哪些研究和资源是必读的?最近,Kyubyong Park 为我们整理了一份完整列表。 GitHub 项目链接:http://ift.tt/2yfj72h 本人从事自然语言处理任务(NLP)的研究已经有很长时间了,有一天我想到,我需要为庞大的 NLP 领域做一个概览,我知道自己肯定不是想要一睹 NLP 任务的全貌的第一个人。 我曾竭尽所能的研究过尽可能多种类型的 NLP 任务,但由于个人知识的局限,我承认还远远没有穷尽整个领域。目前,该项目选取的参考文献都偏重最新的深度学习研究成果。我希望这些能为想要深入钻研一个 NLP 任务的人们提供一个开端。这个项目将持续更新,不过,我更希望与更多人合作。如果你有意愿的话,欢迎对这个项目作出贡献。 回指解析 - See Coreference Resolution (http://ift.tt/2xEDQrO
自动作文评分 - 论文:AutomaticText Scoring Using Neural Networks (http://ift.tt/2yn06Kp
- 论文:ANeural Approach to Automated Essay Scoring (http://ift.tt/2xFfmPm
- 竞赛:Kaggle:The Hewlett Foundation: Automated Essay Scoring (http://ift.tt/2ynjS8y
- 项目:EnhancedAI Scoring Engine(http://ift.tt/2xErpMN
自动语音识别 - WIKI Speech recognition(http://ift.tt/2ymCJ3E
- 论文:DeepSpeech 2: End-to-End Speech Recognition in English and Mandarin (http://ift.tt/2xEEjdF
- 论文:WaveNet:A Generative Model for Raw Audio (http://ift.tt/2ynlFL9
- 项目:A TensorFlow implementation of Baidu's DeepSpeech architecture (http://ift.tt/2xFflLi
- 项目:Speech-to-Text-WaveNet: End-to-end sentence level English speech recognition using DeepMind's WaveNet(http://ift.tt/2ynanGK
- 竞赛:The 5thCHiME Speech Separation and Recognition Challenge (http://ift.tt/2xERcEC
- 资源:The 5thCHiME Speech Separation and Recognition Challenge (http://ift.tt/2ynUsrD
- 资源:CSTRVCTK Corpus (http://ift.tt/2xEEkhJ
- 资源:LibriSpeechASR corpus (http://ift.tt/2ynl4ZM
- 资源:Switchboard-1Telephone Speech Corpus (http://ift.tt/2xEe7ju
- 资源:TED-LIUMCorpus (http://ift.tt/2ynapym
自动摘要 - WIKI Automatic summarization (http://ift.tt/2xEMuqr
- 书籍:AutomaticText Summarization (http://ift.tt/2ynqUKQ
- 论文:TextSummarization Using Neural Networks (http://ift.tt/2xEy9KM
- 论文:Rankingwith Recursive Neural Networks and Its Application to Multi-DocumentSummarization (http://ift.tt/2ynjU0a
- 资源:TextAnalytics Conferences(TAC)(http://ift.tt/2xEMtmn
- 资源:DocumentUnderstanding Conferences (DUC)(http://ift.tt/2ynjSWt
指代消解 - INFO Coreference Resolution(http://ift.tt/2xEwiWd
- 论文:DeepReinforcement Learning for Mention-Ranking Coreference Models (http://ift.tt/2ymMwa0
- 论文:ImprovingCoreference Resolution by Learning Entity-Level Distributed Representations(http://ift.tt/2xERdIG
- 竞赛:CoNLL2012 Shared Task: Modeling Multilingual Unrestricted Coreference in OntoNotes(http://ift.tt/2yn47OV
- 竞赛:CoNLL2011 Shared Task: Modeling Unrestricted Coreference in OntoNotes (http://ift.tt/2xEJM42
实体链接 语法错误纠正 - 论文:NeuralNetwork Translation Models for Grammatical Error Correction (http://ift.tt/2ynqT9K
- 竞赛:CoNLL-2013Shared Task: Grammatical Error Correction (http://ift.tt/2xErLCX
- 竞赛:CoNLL-2014Shared Task: Grammatical Error Correction (http://ift.tt/2yn5UDK
- 资源:NUSNon-commercial research/trial corpus license (http://ift.tt/2xEDvFT
- 资源:Lang-8Learner Corpora(http://ift.tt/2ynj7fG
- 资源:CornellMovie--Dialogs Corpus (http://ift.tt/2xEfvT4
- 项目:DeepText Corrector(http://ift.tt/2ynlG1F
- 产品:deepgrammar(http://ift.tt/2xEwgxo
字素音素转换 - 论文:Grapheme-to-PhonemeModels for (Almost) Any Language (http://ift.tt/2ymCJka
- 论文:PolyglotNeural Language Models: A Case Study in Cross-Lingual Phonetic RepresentationLearning (http://ift.tt/2xEJLgu
- 论文:MultitaskSequence-to-Sequence Models for Grapheme-to-Phoneme Conversion (http://ift.tt/2ymRkfq
- 项目:Sequence-to-Sequence G2P toolkit (http://ift.tt/2xEx9pX
- 资源:Multilingual Pronunciation Data (http://ift.tt/2ynfQNA
语种猜测 语种辨别 - WIKI Language identification (http://ift.tt/2xEfwX8
- 论文:AUTOMATICLANGUAGE IDENTIFICATION USING DEEP NEURAL NETWORKS (http://ift.tt/2ynhO0r
- 竞赛: 2015Language Recognition Evaluation (http://ift.tt/2xEPfYB
语言建模 - WIKI Language model (http://ift.tt/2ymMxL6
- 工具包: KenLMLanguage Model Toolkit (http://ift.tt/2xEe7A0
- 论文:DistributedRepresentations of Words and Phrases and their Compositionality (http://ift.tt/2ynkpaM
- 论文:Character-AwareNeural Language Models (http://ift.tt/2xEMtCT
- 资源: PennTreebank (http://ift.tt/2ymCLbM
语种识别 同一词类 - WIKI Lemmatisation (http://ift.tt/2xEcjHg
- 论文: JointLemmatization and Morphological Tagging with LEMMING (http://ift.tt/2ynkqLS
- 工具包:WordNet Lemmatizer (http://ift.tt/2xErrnT
- 资源:Treebank-3 (http://ift.tt/2ynl6AS
观唇辨意 - WIKI Lip reading (http://ift.tt/2xEQ0Rz
- 论文:LipReading Sentences in the Wild (http://ift.tt/2ynfLcK
- 论文:3DConvolutional Neural Networks for Cross Audio-Visual Matching Recognition (http://ift.tt/2xEwjcJ
- 项目: LipReading - Cross Audio-Visual Recognition using 3D Convolutional Neural Networks(http://ift.tt/2ymTY4Z
- 资源: TheGRID audiovisual sentence corpus (http://ift.tt/2xEPZNv
机器翻译 - 论文:NeuralMachine Translation by Jointly Learning to Align and Translate (http://ift.tt/2ynlHTh
- 论文:NeuralMachine Translation in Linear Time (http://ift.tt/2xEEQfB
- 论文:AttentionIs All You Need (http://ift.tt/2sUoTUo
- 竞赛: ACL2014 NINTH WORKSHOP ON STATISTICAL MACHINE TRANSLATION (http://ift.tt/2ynaq5o
- 竞赛: EMNLP2017 SECOND CONFERENCE ON MACHINE TRANSLATION (WMT17)(http://ift.tt/2xEH52A
- 资源:OpenSubtitles2016 (http://ift.tt/2ynj9nO
- 资源: WIT3:Web Inventory of Transcribed and Translated Talks (https://wit3.fbk.eu/)
- 资源: TheQCRI Educational Domain (QED) Corpus (http://ift.tt/2xEsGU0
生成词法变化 - WIKI Inflection (http://ift.tt/2ynj7wc
- 论文:MorphologicalInflection Generation Using Character Sequence to Sequence Learning (http://ift.tt/2xEqKe4
- 竞赛:SIGMORPHON 2016 Shared Task: Morphological Reinflection (http://ift.tt/2ynnmrR
- 资源:sigmorphon2016 (http://ift.tt/2xErqjP
命名实体消歧 - WIKI Entity linking (http://ift.tt/2ynbF4w
- 论文:Robustand Collective Entity Disambiguation through Semantic Embeddings (http://ift.tt/2xEwkgN
命名实体识别 - WIKI Named-entity recognition (http://ift.tt/2ynlKhV
- 论文:NeuralArchitectures for Named Entity Recognition (http://ift.tt/2xEgKSr
- 项目: OSUTwitter NLP Tools (http://ift.tt/2ynfR46
- 竞赛: NamedEntity Recognition in Twitter (http://ift.tt/2xErqAl
- 竞赛: CoNLL2002 Language-Independent Named Entity Recognition (http://ift.tt/2ymMy1C
- 竞赛:Introduction to the CoNLL-2003 Shared Task: Language-Independent Named EntityRecognition (http://ift.tt/2xEDPUM
- 资源:CoNLL-2002 NER corpus (http://ift.tt/2yn3JAf
- 资源:CoNLL-2003 NER corpus (http://ift.tt/2xEqLi8
- 资源: NUTNamed Entity Recognition in Twitter Shared task (http://ift.tt/2ynksU0
释义检测 - 论文:DynamicPooling and Unfolding Recursive Autoencoders for Paraphrase Detection (http://ift.tt/2xEaJ8f
- 项目:Paralex: Paraphrase-Driven Learning for Open Question Answering (http://ift.tt/2yno0FT
- 资源:Microsoft Research Paraphrase Corpus (http://ift.tt/2xErSi6
- 资源:Microsoft Research Video Description Corpus (http://ift.tt/2yn0993
- 资源: PascalDataset (http://ift.tt/2xEcMsJ
- 资源:Flicker Dataset (http://ift.tt/2ymCLsi
- 资源: TheSICK data set (http://ift.tt/2xErrEp
- 资源: PPDB:The Paraphrase Database (http://ift.tt/2ynl6Ro
- 资源:WikiAnswers Paraphrase Corpus (http://ift.tt/2xEEQw7
语法分析 - WIKI Parsing (http://ift.tt/2ymCNjU
- 工具包: TheStanford Parser: A statistical parser (http://ift.tt/2xEH5j6
- 工具包: spaCyparser (http://ift.tt/2ynjUxc
- 论文:A fastand accurate dependency parser using neural networks (http://ift.tt/2xEJLx0
- 竞赛: CoNLL2017 Shared Task: Multilingual Parsing from Raw Text to Universal Dependencies (http://ift.tt/2ynjTtv
- 竞赛: CoNLL2016 Shared Task: Multilingual Shallow Discourse Parsing (http://ift.tt/2xEsIeA
- 竞赛: CoNLL2015 Shared Task: Shallow Discourse Parsing (http://ift.tt/2ynnkjB
- 竞赛:SemEval-2016 Task 8: The meaning representations may be abstract, but this taskis concrete! (http://ift.tt/2xEgK4T
词性标记 - WIKI Part-of-speech tagging (http://ift.tt/2yneGSd
- 论文:MultilingualPart-of-Speech Tagging with Bidirectional Long Short-Term Memory Models andAuxiliary Loss (http://ift.tt/2xERdbE
- 论文:UnsupervisedPart-Of-Speech Tagging with Anchor Hidden Markov Models (http://ift.tt/2yn5YU0
- 资源:Treebank-3 (http://ift.tt/2ynl6AS
- 工具包:nltk.tag package (http://ift.tt/2xEwkxj
拼音-中文转换 - 论文:NeuralNetwork Language Model for Chinese Pinyin Input Method Engine (http://ift.tt/2ynfSVI
- 项目: NeuralChinese Transliterator (http://ift.tt/2xEgL8X
问答系统 - WIKI Question answering (http://ift.tt/2ynovjc
- 论文:Ask MeAnything: Dynamic Memory Networks for Natural Language Processing (http://ift.tt/2xEx9Gt
- 论文:DynamicMemory Networks for Visual and Textual Question Answering (http://ift.tt/2ynbFl2
- 竞赛: TRECQuestion Answering Task (http://ift.tt/2xEyahO
- 竞赛:NTCIR-8: Advanced Cross-lingual Information Access (ACLIA)(http://ift.tt/2ynniIv
- 竞赛: CLEFQuestion Answering Track (http://ift.tt/2xEqKuA
- 竞赛:SemEval-2017 Task 3: Community Question Answering (http://ift.tt/2ynhQp5
- 资源: MSMARCO: Microsoft MAchine Reading COmprehension Dataset (http://ift.tt/2xErM9Z
- 资源:Maluuba NewsQA (http://ift.tt/2ynhRt9
- 资源: SQuAD:100,000+ Questions for Machine Comprehension of Text (http://ift.tt/2xExaKx
- 资源:GraphQuestions: A Characteristic-rich Question Answering Dataset (http://ift.tt/2ynhS0b
- 资源: StoryCloze Test and ROCStories Corpora (http://ift.tt/2xEe9b6
- 资源:Microsoft Research WikiQA Corpus (http://ift.tt/2ynfTsK
- 资源:DeepMind Q&A Dataset (http://ift.tt/2xEckei
- 资源: QASent(http://ift.tt/2ynfTZM
关系提取 - WIKI Relationship extraction (http://ift.tt/2xEPgvD
- 论文:A deeplearning approach for relationship extraction from interaction context insocial manufacturing paradigm (http://ift.tt/2ynhSxd
语义角色标注 - WIKI Semantic role labeling (http://ift.tt/2xEPhzH
- 书籍:Semantic Role Labeling (http://ift.tt/2ynhT4f
- 论文:End-to-endLearning of Semantic Role Labeling Using Recurrent Neural Networks (http://ift.tt/2xEe872
- 论文:NeuralSemantic Role Labeling with Dependency Path Embeddi ngs (http://ift.tt/2ynfUwO
- 竞赛:CoNLL-2005 Shared Task: Semantic Role Labeling (http://ift.tt/2xFfnCU
- 竞赛:CoNLL-2004 Shared Task: Semantic Role Labeling (http://ift.tt/2ynhU8j
- 工具包:Illinois Semantic Role Labeler (SRL)(http://ift.tt/2xEEl5h
- 资源:CoNLL-2005 Shared Task: Semantic Role Labeling (http://ift.tt/2ynfV3Q
语句边界消歧 - WIKI Sentence boundary disambiguation (http://ift.tt/2xEDRfm
- 论文:AQuantitative and Qualitative Evaluation of Sentence Boundary Detection for theClinical Domain (http://ift.tt/2ynhUFl
- 工具包: NLTKTokenizers (http://ift.tt/2xEgLpt
- 资源: TheBritish National Corpus (http://ift.tt/2ynfVAS
- 资源:Switchboard-1 Telephone Speech Corpus (http://ift.tt/2xEe7ju
情绪分析 - WIKI Sentiment analysis (http://ift.tt/2xEEPIz
- INFO Awesome Sentiment Analysis (http://ift.tt/2ynfW7U
- 竞赛:Kaggle: UMICH SI650 - Sentiment Classification (http://ift.tt/2xEe9rC
- 竞赛:SemEval-2017 Task 4: Sentiment Analysis in Twitter (http://ift.tt/2ynfWEW
- 竞赛:SemEval-2017 Task 5: Fine-Grained Sentiment Analysis on Financial Microblogsand News (http://ift.tt/2xErSP8
- 项目:SenticNet (http://ift.tt/2ynhVcn
- 资源:Multi-Domain Sentiment Dataset (version2.0)(http://ift.tt/2xEgKlp
- 资源:Stanford Sentiment Treebank (http://ift.tt/2ynhVJp
- 资源:Twitter Sentiment Corpus (http://ift.tt/2xEckuO
- 资源:Twitter Sentiment Analysis Training Corpus (http://ift.tt/2ynhWgr
- 资源: AFINN:List of English words rated for valence (http://ift.tt/2xEwjJL
源分离 - WIKI Source separation (http://ift.tt/2ynfXbY
- 论文:FromBlind to Guided Audio Source Separation (http://ift.tt/2xEsHr2
- 论文:JointOptimization of Masks and Deep Recurrent Neural Networks for Monaural SourceSeparation (http://ift.tt/2ynl77U
- 竞赛: SignalSeparation Evaluation Campaign (SiSEC)(http://ift.tt/2xEcMZL
- 竞赛: CHiMESpeech Separation and Recognition Challenge (http://ift.tt/2xERcEC
说话人认证 语音身份分离 - WIKI Speaker diarisation (http://ift.tt/2ynfXJ0
- 论文:DNN-basedspeaker clustering for speaker diarisation (http://ift.tt/2xEwkNP
- 论文:UnsupervisedMethods for Speaker Diarization: An Integrated and Iterative Approach (http://ift.tt/2ynUv6N
- 论文:Audio-VisualSpeaker Diarization Based on Spatiotemporal Bayesian Fusion (http://ift.tt/2xEyayk
- 竞赛: RichTranscription Evaluation (http://ift.tt/2yneH8J
说话人识别 - WIKI Speaker recognition (http://ift.tt/2xEQ1F7
- 论文:A NOVELSCHEME FOR SPEAKER RECOGNITION USING A PHONETICALLY-AWARE DEEP NEURAL NETWORK (http://ift.tt/2ynfYN4
- 论文:DEEPNEURAL NETWORKS FOR SMALL FOOTPRINT TEXT-DEPENDENT SPEAKER VERIFICATION (http://ift.tt/2xEfxKG
- 竞赛: NISTSpeaker Recognition Evaluation (SRE)(http://ift.tt/2ynj9UQ
- INFO Are there any suggestions for free databases for speakerrecognition? (http://ift.tt/2xEPgM9
唇读 语音识别 语音分割 - WIKI Speech_segmentation (http://ift.tt/2ynovzI
- 论文:WordSegmentation by 8-Month-Olds: When Speech Cues Count More Than Statistics (http://ift.tt/2xFfnTq
- 论文:UnsupervisedWord Segmentation and Lexicon Discovery Using Acoustic Word Embeddings (http://ift.tt/2ynkriU
- 论文:UnsupervisedLexicon Discovery from Acoustic Inpu (http://ift.tt/2xEJMRA
- 论文:Weaklysupervised spoken term discovery using cross-lingual side information (http://ift.tt/2yn5XiU
- 资源:CALLHOME Spanish Speech (http://ift.tt/2xEclyS
语音合成 - WIKI Speech synthesis (http://ift.tt/2ynktaw
- 论文:WaveNet:A Generative Model for Raw Audio (http://ift.tt/2ynlFL9
- 论文:Tacotron:Towards End-to-End Speech Synthesis (http://ift.tt/2xEaKJl
- 论文:DeepVoice 2: Multi-Speaker Neural Text-to-Speech (http://ift.tt/2ynqXX2
- 资源: TheWorld English Bible (http://ift.tt/2xEcO3P
- 资源: LJSpeech Dataset (http://ift.tt/2ynnl7h
- 资源: LessacData (http://ift.tt/2xEDRvS
- 竞赛:Blizzard Challenge 2017 (http://ift.tt/2ymRmE4
- PRODUCT Lyrebird (https://lyrebird.ai/)
- 项目: TheFestvox project (http://ift.tt/2xEQ2Jb
- 工具包:Merlin: The Neural Network (NN) based Speech Synthesis System (http://ift.tt/2ynbHcE
语音增强 - WIKI Speech enhancement (http://ift.tt/2xEDxxv
- 书籍: Speechenhancement: theory and practice (http://ift.tt/2ynjUNI
- 论文 AnExperimental Study on Speech Enhancement BasedonDeepNeuralNetwork (http://ift.tt/2xEEllN
- 论文: ARegression Approach to Speech Enhancement BasedonDeepNeuralNetworks (http://ift.tt/2ynasu2
- 论文: SpeechEnhancement Based on Deep Denoising Autoencoder (http://ift.tt/2xEReMK
语音文本转换 口语的术语检测 词干提取 - WIKI Stemming (http://ift.tt/2ymMyyE
- 论文: ABACKPROPAGATION NEURAL NETWORK TO IMPROVE ARABIC STEMMING (http://ift.tt/2xEckLk
- 工具包: NLTKStemmers (http://ift.tt/2ymCNAq
术语提取 - WIKI Terminology extraction (http://ift.tt/2xEsILC
- 论文: NeuralAttention Models for Sequence Classification: Analysis and Application to KeyTerm Extraction and Dialogue Act Detection (http://ift.tt/2ynUvnj
文本简化 - WIKI Text simplification (http://ift.tt/2xFfoXu
- 论文:Aligning Sentences from Standard Wikipedia to Simple Wikipedia (http://ift.tt/2ynqWlW
- 论文:Problems in Current Text Simplification Research: New Data Can Help (http://ift.tt/2xErNuz
- 资源:Newsela Data (http://ift.tt/2ynhYoz
文本语音转换 文本蕴涵 - WIKI Textual entailment (http://ift.tt/2xEwl4l
- 项目:Textual Entailment with TensorFlow (http://ift.tt/2ynj8jK
- 论文:Textual Entailment with Structured Attentions and Composition (http://ift.tt/2xEES7d
- 竞赛:SemEval-2014 Task 1: Evaluation of compositional distributional semantic modelson full sentences through semantic relatedness and textual entailment (http://ift.tt/2ynnkA7
- 竞赛:SemEval-2013 Task 7: The Joint Student Response Analysis and 8th RecognizingTextual Entailment Challenge (http://ift.tt/2xEDSzW
声音转换 - 论文:PHONETIC POSTERIORGRAMS FOR MANY-TO-ONE VOICE CONVERSION WITHOUT PARALLEL DATATRAINING (http://ift.tt/2ynqYdy
- 项目: Animplementation of voice conversion system utilizing phonetic posteriorgrams (http://ift.tt/2xExbhz
- 竞赛: VoiceConversion Challenge 2016 (http://ift.tt/2ynjabm
- 竞赛: VoiceConversion Challenge 2018 (http://ift.tt/2xEEmpR
- 资源:CMU_ARCTIC speech synthesis databases (http://ift.tt/2ymyPYv
- 资源: TIMITAcoustic-Phonetic Continuous Speech Corpus (http://ift.tt/2xEPinf
声音识别 词嵌入 - WIKI Word embedding (http://ift.tt/2ynkrzq
- 工具包:Gensim: word2vec (http://ift.tt/2xEycWY
- 工具包:fastText (http://ift.tt/2ynnXtH
- 工具包: GloVe:Global Vectors for Word Representation (http://ift.tt/2xEwhBs
- INFO Where to get a pretrained model (http://ift.tt/2ynktr2
- 项目:Pre-trained word vectors of 30+ languages (http://ift.tt/2xEMvL1
- 项目:Polyglot: Distributed word representations for multilingual NLP (http://ift.tt/2ynbHta
词预测 - INFO What is Word Prediction? (http://ift.tt/2xEeaMc
- 论文: Theprediction of character based on recurrent neural network language model (http://ift.tt/2ynnYxL
- 论文: AnEmbedded Deep Learning based Word Prediction (http://ift.tt/2xEJN86
- 论文:Evaluating Word Prediction: Framing Keystroke Savings (http://ift.tt/2ynnZ4N
- 资源: AnEmbedded Deep Learning based Word Prediction (http://ift.tt/2xEH66E
- 项目: WordPrediction using Convolutional Neural Networks—can you do better than iPhone™Keyboard? (http://ift.tt/2ynl9g2
词分割 - WIKI Word segmentation (http://ift.tt/2xEebQg
- 论文: NeuralWord Segmentation Learning for Chinese (http://ift.tt/2yn5Zr2
- 项目:Convolutional neural network for Chinese word segmentation (http://ift.tt/2xEecDO
- 工具包:Stanford Word Segmenter (http://ift.tt/2ymTX0V
- 工具包: NLTKTokenizers (http://ift.tt/2xEgLpt
词义消歧 - 资源:Word-sense disambiguation (http://ift.tt/2xEDxO1
- 论文:Train-O-Matic: Large-Scale Supervised Word Sense Disambiguation in MultipleLanguages without Manual Training Data (http://ift.tt/2ymTYC1
- 资源:Train-O-Matic Data (http://ift.tt/2xEElCj
- 资源:BabelNet (http://babelnet.org/)
]]> 原文: http://ift.tt/2ynjUh3 | | 机器知心 | | |
没有评论:
发表评论