https://DevOpsCloud.io -- Cloud Monk Losang Jinpa

Return to Practical Natural Language Processing, Python NLP, NLP, NLP bibliography, Python AI - AI bibliography, Python DL - Machine Learning (ML) bibliography, Python ML - Deep Learning (DL) bibliography, Python Data science - Data Science bibliography

“ (PrctNLP 2020)

Index

• A/B testing, Practical Advice, Image match, Iterating Existing Models
• Abstraction and Reasoning Corpus (ARC) dataset, Peeking over the Horizon, Peeking over the Horizon
• abstractive summarization, Text Summarization, Practical Advice
• accounting and auditing, Accounting and auditing
• accuracy, Intrinsic Evaluation
• ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD), The KDD Process
• active learning, Data Acquisition
• NER using, NER Using Active Learning-NER Using Active Learning
• with Prodigy, Case Study: Corporate Ticketing
• text classification, Less Training Data: Active Learning and Domain Adaptation-Less Training Data: Active Learning and Domain Adaptation, Case Study: Corporate Ticketing
• Adamic, Lada, Identifying Memes
• adapting to new domains, Learning with No or Less Data and Adapting to New Domains-Less Training Data: Active Learning and Domain Adaptation
• adult content filtering, Applications
• advanced processing, Pre-Processing, Advanced Processing-Advanced Processing
• AI (see artificial intelligence)
• Airbnb, Practical Advice, Product Search
• Airflow (Apache), Revisiting the NLP Pipeline: Deploying NLP Software
• Algolia, A Case Study: Book Store Search
• Allen AI, Peeking over the Horizon
• AllenNLP, NER Using an Existing Library, Approaches to RE
• DeepQA library, Looking for Deeper Answers
• Grover, Fake News
• Amazon, NLP in the Real World, Applications, Topics in Brief
• faceted search on, Search in E-Commerce
• “Reviews that mention” filter, Keyphrase Extraction
• sentiment analysis APIs, Using Existing Text Classification APIs
• Amazon Alexa, NLP: A Primer, NLP in the Real World, NLP Tasks, An NLP Walkthrough: Conversational Agents, Chatbots
• Amazon Comprehend, Building Your Model, Case Study
• Amazon Comprehend Medical, Medical Information Extraction and Analysis-Medical Information Extraction and Analysis
• Amazon Mechanical Turk, No Training Data
• Amazon Research, Fake News
• Amazon Translate, NLP in the Real World
• Amazon Web Services (AWS), Deployment
• AWS Cloud, Revisiting the NLP Pipeline: Deploying NLP Software
• AWS S3, Revisiting the NLP Pipeline: Deploying NLP Software
• SageMaker, Revisiting the NLP Pipeline: Deploying NLP Software, An Example Scenario
• ambiguity , Ambiguity-Ambiguity
• anomaly detection, Monitoring
• answer extraction, Question-Answering Systems
• answering questions (see question answering)
• Apache Airflow, Revisiting the NLP Pipeline: Deploying NLP Software
• Apache Nutch, Components of a Search Engine
• Apache Oozie, Revisiting the NLP Pipeline: Deploying NLP Software
• Apache Solr software, Components of a Search Engine
• APIs
• integration of, Rasa NLU
• modeling with, Building Your Model
• MT, Using a Machine Translation API: An Example-Using a Machine Translation API: An Example
• product categorization and taxonomy, Product Categorization and Taxonomy, Product Categorization and Taxonomy
• text classification with, Using Existing Text Classification APIs, Case Study: Corporate Ticketing
• Apple, Customer Support on Social Channels
• Apple Siri, NLP: A Primer, NLP in the Real World, NLP Tasks, An NLP Walkthrough: Conversational Agents
• Arabic, Working with Other Languages
• ARC (see Abstraction and Reasoning Corpus)
• Arria, NLP in the Real World
• artificial intelligence (AI), Machine Learning, Deep Learning, and NLP: An Overview, Chatbots, The End-to-End NLP Process
• expectations vs reality, Right Problem and Right Expectations
• further reading, Other Aspects-Other Aspects
• key points and rules of thumb, Making AI Succeed at Your Organization-Other Aspects
• perspectives on, Peeking over the Horizon
• for social good, Peeking over the Horizon
• aspect-based sentiment analysis, Applications-Applications, Aspect-Level Sentiment Analysis-Unsupervised approach
• aspects, Aspect-Level Sentiment Analysis
• connecting overall ratings to, Connecting Overall Ratings to Aspects-Connecting Overall Ratings to Aspects
• understanding, Understanding Aspects-Understanding Aspects
• ASR (automatic speech recognition), System-Specific Error Correction
• assessment tools, NLP in the Real World
• ATIS (Airline Travel Information Systems) dataset, Datasets
• attention, Outcome prediction and best practices
• attention networks, Troubleshooting and Interpretability
• attribute extraction, E-Commerce Catalog, Attribute Extraction-Indirect attribute extraction
• algorithms, Attribute Extraction
• derived, Attribute Extraction
• direct, Attribute Extraction, Direct attribute extraction-Direct attribute extraction
• indirect, Indirect attribute extraction-Indirect attribute extraction
• latent, Latent attribute extraction from reviews
• attribute match, Attribute match
• AUC, Intrinsic Evaluation
• Audit Command Language (Deloitte), Accounting and auditing
• auditing, Accounting and auditing
• authorship attribution, Applications
• AutoCompete framework, Google Cloud AutoML and other techniques
• autoencoders, Autoencoders-Autoencoders
• automated machine learning (AutoML), Automating Machine Learning-Google Cloud AutoML and other techniques
• automated scoring, NLP in the Real World, Handcrafted Feature Representations
• Automatic Content Extraction Program, Information Extraction
• automatic response generation, Response Generation
• automatic speech recognition (ASR), System-Specific Error Correction
• automation, full, Other Aspects
• autosklearn, auto-sklearn-auto-sklearn
• autotldr bot (Reddit), Summarization Use Cases
• AWS (see Amazon Web Services)

• back translation, Data Acquisition, Data Acquisition, Practical Advice
• bag of n-grams (BoN), Bag of N-Grams-Bag of N-Grams
• bag of words (BoW), Bag of Words-Bag of Words, Text Representation for SMTD, Fake News
• baselines, A Good Process
• Bayes’ theorem, Naive Bayes
• BBC, Template Filling
• BCG, Peeking over the Horizon
• Beautiful Soup library, HTML Parsing and Cleanup, Removing markup elements
• Bender, Emily, Context
• BERT (Bidirectional Encoder Representations from Transformers), Transformers, Universal Text Representations, Models
• fact-verification with, Fake News
• search with, A Typical Enterprise Search Pipeline
• size, An Example Scenario
• text classification with, Text Classification with Large, Pre-Trained Language Models-Text Classification with Large, Pre-Trained Language Models
• BERT for Biomedical Text (BioBERT), Medical Information Extraction and Analysis, Medical Information Extraction and Analysis
• BERT for financial text (FinBERT), Financial sentiment
• best practices, Outcome prediction and best practices
• bias, Universal Text Representations, Practical Advice
• bigram flipping, Data Acquisition
• billing, Patient Prioritization and Billing
• binary classification, Text Classification
• Bing (search engine), NLP in the Real World, Question-Answering Systems
• Bing Answer Search API, Looking for Deeper Answers
• Bing Microsoft Translator, NLP in the Real World
• Bing Translate API, Using a Machine Translation API: An Example
• BioBERT (BERT for Biomedical Text), Medical Information Extraction and Analysis, Medical Information Extraction and Analysis
• biographical information extraction, Search and Information Retrieval
• BLEU (Bilingual Evaluation Understudy), Intrinsic Evaluation, Intrinsic Evaluation
• Bloomberg, Financial sentiment
• Bloomberg Terminal, Event Extraction
• BoN (bag of n-grams), Bag of N-Grams-Bag of N-Grams
• bootstrapping, No Training Data
• BoW (bag of words), Bag of Words-Bag of Words, Text Representation for SMTD, Fake News
• Buoy system, Health Assistants
• Burmese, Working with Other Languages
• business issues, Peeking over the Horizon

• Capital Float and Microbnk, Risk assessments
• Captcha tests, No Training Data
• Carnegie Mellon University, Minimizing Technical Debt
• Casari, Amanda, Minimizing Technical Debt
• catalogs, E-Commerce Catalog-Product Recommendations, Building an E-Commerce Catalog-Image match
• categorization, product, Product Categorization and Taxonomy-Product Categorization and Taxonomy
• CBOW (continuous bag of words), CBOW-CBOW, SkipGram, Going Beyond Words
• CDQA-Suite library, Looking for Deeper Answers
• CFG (context-free grammar), Heuristics-Based NLP
• character embeddings, Text Representation for SMTD
• character models, Mental Healthcare Monitoring
• character n-gram embeddings, Text Representation for SMTD
• character n-gram–based models, Mental Healthcare Monitoring
• characteristic features, Direct attribute extraction
• chatbots, IE Applications, Chatbots-Wrapping Up, Other Aspects
• applications, Applications-A Simple FAQ Bot
• build walkthrough, PizzaStop Chatbot-Testing our agent
• case study, A Case Study: Recipe Recommendations-Open-Ended Generative Chatbots
• chitchat types, A Taxonomy of Chatbots
• exact answer, A Taxonomy of Chatbots
• examples, Chatbots
• existing frameworks, Utilizing Existing Frameworks-Utilizing Existing Frameworks
• FAQ bot, A Simple FAQ Bot-A Taxonomy of Chatbots, A Taxonomy of Chatbots
• flow-based, A Taxonomy of Chatbots
• goal-oriented dialog, Goal-Oriented Dialog
• healthcare, Health Assistants-Health Assistants
• open, A Taxonomy of Chatbots
• open-ended, A Taxonomy of Chatbots, Open-Ended Generative Chatbots-Open-Ended Generative Chatbots
• pipeline for, A Pipeline for Building Dialog Systems-A Pipeline for Building Dialog Systems
• taxonomy of, A Taxonomy of Chatbots-Chitchats
• terminology, Dialog Systems in Detail
• testing, Testing our agent-Testing our agent
• text-based, A Pipeline for Building Dialog Systems
• voice-based, A Pipeline for Building Dialog Systems
• Chee, Cedric, An Example Scenario
• Chef, Revisiting the NLP Pipeline: Deploying NLP Software
• Chinese, Working with Other Languages
• chitchats, A Taxonomy of Chatbots
• Chollet, François, Peeking over the Horizon
• Chomsky, Noam, NLP: A Primer, Heuristics-Based NLP
• Chronic, Temporal Information Extraction
• CJK languages, Working with Other Languages
• classification, Text Classification
• (see also text classification)
• binary, Text Classification
• multiclass, Text Classification
• product categorization and taxonomy, Product Categorization and Taxonomy-Product Categorization and Taxonomy
• topic, Text Classification
• classification tasks, Machine Learning for NLP, Intrinsic Evaluation
• clinical decision support systems, Clinical Decision Support Systems
• clinical information systems, HARVEST: Longitudinal report understanding-HARVEST: Longitudinal report understanding
• cloud storage, Revisiting the NLP Pipeline: Deploying NLP Software
• clpsych.org, Applications
• CMU Book Summaries dataset, Setting Up a Search Engine: An Example
• CNNs (convolutional neural networks), Convolutional neural networks, Deep Learning for Text Classification, CNNs for Text Classification-CNNs for Text Classification, Models
• code mixing, Code mixing and transliteration
• code reproducibility, Code and Model Reproducibility
• coherence, Training a Topic Model: An Example
• collaborative filtering, Recommender Systems for Textual Data
• Columbia University, HARVEST: Longitudinal report understanding
• common knowledge, Common knowledge-Common knowledge
• common sense, Why Deep Learning Is Not Yet the Silver Bullet for NLP
• complements, A Case Study: Substitutes and Complements, A Case Study: Substitutes and Complements
• compute costs, Other Aspects
• conditional random fields (CRFs), Conditional random fields, Building an NER System
• confusion matrix, Intrinsic Evaluation
• congratsbot, Event Extraction
• CONLL-03 dataset, Building an NER System
• connotation, Distributed Representations
• content acquisition, A Typical Enterprise Search Pipeline
• content classification and organization, Applications
• content discovery
• chatbots in, Applications
• with emotional content, Mental Healthcare Monitoring
• legal, NLP and the Legal Landscape
• context, What Is Language?, Context
• context window, SkipGram
• context-based conversations, Rasa NLU
• context-free grammar (CFG), Heuristics-Based NLP
• contextual features, Direct attribute extraction
• contextual word representations, Universal Text Representations
• continuous bag of words (CBOW), CBOW-CBOW, SkipGram, Going Beyond Words
• contract generation, NLP and the Legal Landscape
• contract review, NLP and the Legal Landscape
• ContraxSuite, Legal entity extraction with LexNLP, Legal entity extraction with LexNLP
• conversational agents, NLP Tasks
• case study, An NLP Walkthrough: Conversational Agents-An NLP Walkthrough: Conversational Agents
• typical flow, An NLP Walkthrough: Conversational Agents
• conversations
• actionable, Customer Support on Social Channels
• context-based, Rasa NLU
• convolutional neural networks (CNNs), Deep Learning for Text Classification, CNNs for Text Classification-CNNs for Text Classification, Models
• Coppersmith, Glen, Mental Healthcare Monitoring
• coreference resolution, An NLP Walkthrough: Conversational Agents, Advanced Processing
• corporate ticketing (case study), Case Study: Corporate Ticketing-Case Study: Corporate Ticketing
• cost of compute, Other Aspects
• costs, Why Deep Learning Is Not Yet the Silver Bullet for NLP
• COTA (Customer Obsession Ticketing Assistant), Case Study-Case Study
• covariate shift, Building and Maintaining a Mature System
• crawlers, Components of a Search Engine
• crawling/content acquisition, A Typical Enterprise Search Pipeline
• creativity, Creativity
• credit risk, Risk assessments-Risk assessments
• CRFs (conditional random fields), Conditional random fields, Building an NER System
• CRM (customer relationship management) , Customer Support on Social Channels
• crowdsourcing, No Training Data
• Customer Obsession Ticketing Assistant (COTA), Case Study-Case Study
• customer relationship management (CRM), Customer Support on Social Channels
• customer support
• chatbots in, Applications
• on social channels, Applications, Applications, Customer Support on Social Channels-Memes and Fake News

• DARPA, Chatbots
• Dartmouth College, Machine Learning, Deep Learning, and NLP: An Overview
• Das, Vir, Social Media
• data acquisition, NLP Pipeline-Data Acquisition
• data annotation, Rasa NLU-Rasa NLU
• data augmentation, Data Acquisition-Data Acquisition
• data extraction (see information extraction)
• data labeling, Data and Timing
• data mining, The KDD Process
• data pre-processing, The KDD Process
• (see also pre-processing)
• data reduction, The KDD Process
• data science (DS), The End-to-End NLP Process, The Data Science Process-Microsoft Team Data Science Process
• Data Version Control (DVC), Code and Model Reproducibility
• databases, in-memory, Universal Text Representations
• Dataminr, Financial sentiment
• datasets
• complex, Question answering for health
• crowdsourcing for, No Training Data
• for dialog systems, Datasets, Wrapping Up
• fact-verification with, Fake News
• goal-oriented, Datasets
• i2b2, Question answering for health
• public, Data Acquisition, Case Study: Corporate Ticketing, Understanding Twitter Sentiment
• target, The KDD Process
• text classification with, One Pipeline, Many Classifiers-One Pipeline, Many Classifiers, Case Study: Corporate Ticketing
• version control , Code and Model Reproducibility
• Dawkins, Richard, Identifying Memes
• DBOW (distributed bag of words), Distributed Representations Beyond Words and Characters
• DBpedia dataset, Subword Embeddings and fastText
• DBpedia Spotlight, NEL Using Azure API
• decision support systems, Clinical Decision Support Systems
• deep learning (DL)
• attribute extraction with, Direct attribute extraction
• for dialogue generation, Deep Reinforcement Learning for Dialogue Generation-Human-in-the-Loop
• fact-verification with, Fake News
• feature engineering, Feature Engineering, DL Pipeline
• limitations, Why Deep Learning Is Not Yet the Silver Bullet for NLP-Why Deep Learning Is Not Yet the Silver Bullet for NLP
• for NLP, Deep Learning for NLP-Autoencoders
• overview, Machine Learning, Deep Learning, and NLP: An Overview-Machine Learning, Deep Learning, and NLP: An Overview
• question answering with, Looking for Deeper Answers
• reinforcement learning, Deep Reinforcement Learning for Dialogue Generation-Human-in-the-Loop
• for text classification, Deep Learning for Text Classification-Text Classification with Large, Pre-Trained Language Models
• with EHRs, Outcome prediction and best practices
• deep neural networks, Looking for Deeper Answers
• DeepQA library, Looking for Deeper Answers
• Deloitte, Accounting and auditing
• Demoji, Handling emojis
• denotation, Distributed Representations
• deployment, Deployment, Revisiting the NLP Pipeline: Deploying NLP Software-Revisiting the NLP Pipeline: Deploying NLP Software
• device, Why Deep Learning Is Not Yet the Silver Bullet for NLP
• example scenario, An Example Scenario-An Example Scenario
• typical steps, Revisiting the NLP Pipeline: Deploying NLP Software
• derived attribute extraction, Attribute Extraction
• deterministic matches, Heuristics-Based NLP
• device deployment, Why Deep Learning Is Not Yet the Silver Bullet for NLP
• diagnosis chatbots, Health Assistants, Health Assistants
• dialog act classification, Dialog Act Classification
• dialog act or intent, Dialog Systems in Detail
• dialog act prediction, Dialog act prediction-Models
• dialog management, An NLP Walkthrough: Conversational Agents
• dialog managers, A Pipeline for Building Dialog Systems
• dialog state or context, Dialog Systems in Detail
• dialog systems
• components, Deep Dive into Components of a Dialog System-Models
• datasets for, Datasets, Wrapping Up
• deep reinforcement learning for, Deep Reinforcement Learning for Dialogue Generation-Human-in-the-Loop
• in detail, Dialog Systems in Detail-Testing our agent
• end-to-end approach, End-to-End Approach
• examples with code walkthrough, Dialog Examples with Code Walkthrough-Models
• human-in-the-loop, Human-in-the-Loop-Human-in-the-Loop
• other pipelines, Other Dialog Pipelines-Human-in-the-Loop
• pipeline for, A Pipeline for Building Dialog Systems-A Pipeline for Building Dialog Systems
• terminology, Dialog Systems in Detail
• Dialogflow (Google), PizzaStop Chatbot, Utilizing Existing Frameworks
• chatbot build with, PizzaStop Chatbot-Testing our agent
• creating agents with, Building our Dialogflow agent
• digits: removing, Frequent Steps
• dimensionality, Deep Learning for Text Classification
• direct attribute extraction, Attribute Extraction, Direct attribute extraction-Direct attribute extraction
• discriminative classifier, Logistic Regression
• displaCy visualizer, Named Entity Recognition
• distant supervision, Approaches to RE
• distributed bag of words (DBOW), Distributed Representations Beyond Words and Characters
• distributed memory (DM), Distributed Representations Beyond Words and Characters
• distributed representations, Distributed Representations-Distributed Representations Beyond Words and Characters
• distributional hypothesis, Distributed Representations
• distributional similarity, Distributed Representations
• diversity across languages, Diversity across languages
• Django Project, An Example Scenario
• DL (see deep learning)
• DM (distributed memory), Distributed Representations Beyond Words and Characters
• Doc2vec model, Distributed Representations Beyond Words and Characters, Document Embeddings
• serving recommendations with, Creating a Book Recommender System: An Example
• text classification with, Document Embeddings-Document Embeddings
• training, Document Embeddings
• Docker, An Example Scenario
• document categorization, Text Classification
• (see also text classification)
• document embeddings
• text classification with, Document Embeddings-Document Embeddings
• visualizing, Visualizing Embeddings-Visualizing Embeddings
• Document Understanding Conference series, Text Summarization
• domain adaptation, Less Training Data: Active Learning and Domain Adaptation-Less Training Data: Active Learning and Domain Adaptation
• Domingoes, Pedro, Minimizing Technical Debt
• DS (data science), The End-to-End NLP Process, The Data Science Process-Microsoft Team Data Science Process
• DSTC dataset, Datasets
• Duckling library, Temporal Information Extraction
• Duolingo, NLP in the Real World

• e-commerce and retail, E-Commerce and Retail-Wrapping Up
• catalogs, E-Commerce Catalog-Product Recommendations, Building an E-Commerce Catalog-Image match
• chatbots, Applications
• NLP applications, NLP in the Real World, E-Commerce and Retail
• product categorization and taxonomy, Product Categorization and Taxonomy-Product Categorization and Taxonomy
• product deduplication and matching, Product Deduplication and Matching-Image match
• product enrichment, Product Enrichment-Product Enrichment
• recommendations for, Recommendations for E-Commerce-Product linking
• review analysis, Review Analysis-Understanding Aspects
• search, Search in E-Commerce-Search in E-Commerce
• text classification, Applications-Applications
• Earley parser, Heuristics-Based NLP
• Easy Data Augmentation (EDA), Data Acquisition
• eBay, Applications, Product Categorization and Taxonomy
• Economic News Article Tone and Relevance dataset (Figure Eight), One Pipeline, Many Classifiers, Explaining Classifier Predictions with Lime
• EDA (Easy Data Augmentation), Data Acquisition
• editorial reviews, Understanding Aspects
• Educational Testing Service (ETS), Handcrafted Feature Representations, Handcrafted Feature Representations
• EHRs (see electronic health records)
• Elastic on Azure, A Case Study: Book Store Search
• Elasticsearch, Setting Up a Search Engine: An Example-Setting Up a Search Engine: An Example
• DSL, Search in E-Commerce
• ELK stack, Monitoring
• Python API, Setting Up a Search Engine: An Example
• Elasticsearch Learning to Rank, A Case Study: Book Store Search
• electronic health records (EHRs), Electronic Health Records-Outcome prediction and best practices
• example, Health and Medical Records
• Eliza chatbot, Chatbots
• ELK stack (Elastic), Monitoring
• ELMo, Universal Text Representations
• email platforms, NLP in the Real World
• emails: IE from, Temporal Information Extraction, Case Study
• embedding, Distributed Representations
• character, Text Representation for SMTD
• character n-gram, Text Representation for SMTD
• pre-trained, Universal Text Representations, Text Representation for SMTD
• subword, Subword Embeddings and fastText-Subword Embeddings and fastText
• visualizing, Visualizing Embeddings-Visualizing Embeddings
• word, Word Embeddings-SkipGram
• word-based, Text Representation for SMTD
• beyond words, Going Beyond Words-Going Beyond Words
• emojis, Handling emojis
• emotion classification, Mental Healthcare Monitoring, Mental Healthcare Monitoring
• emr, Question answering for health
• emrQA dataset, Question answering for health, Question answering for health
• Enam, S. Zayd, Other Aspects
• English, Phonemes, Recurrent neural networks, Working with Other Languages, Practical Advice
• ensembling, Practical Advice
• enterprise search engines, Search and Information Retrieval, A Typical Enterprise Search Pipeline-A Typical Enterprise Search Pipeline
• entity linking, Named Entity Disambiguation and Linking
• EntityRuler (spaCy), Building an NER System, NER Using Active Learning
• Ernst & Young, Accounting and auditing
• error correction, System-Specific Error Correction-System-Specific Error Correction
• ETS (Educational Testing Service), Handcrafted Feature Representations, Handcrafted Feature Representations
• evaluation, Evaluation-Extrinsic Evaluation, Peeking over the Horizon
• extrinsic, Evaluation, Extrinsic Evaluation-Extrinsic Evaluation
• intrinsic, Intrinsic Evaluation-Intrinsic Evaluation
• visual methods, Intrinsic Evaluation
• event extraction, IE Tasks, Event Extraction-Event Extraction
• extractive summarization, Text Summarization, Practical Advice

• F1 score, Intrinsic Evaluation
• Facebook, Social Media
• customer support on, Customer Support on Social Channels
• data acquisition, Data Acquisition
• fastText embeddings, Pre-trained word embeddings, Going Beyond Words, Subword Embeddings and fastText
• fastText library, Subword Embeddings and fastText-Subword Embeddings and fastText
• memes on, Identifying Memes
• Facebook AI Research, An Example Scenario
• Facebook Messenger, Temporal Information Extraction, Chatbots
• faceted search, Search in E-Commerce-Search in E-Commerce
• facets, Search in E-Commerce
• fact-verification, Fake News
• fake news, Applications, Applications, Memes and Fake News, Fake News-Fake News
• Falcon Web Framework, An Example Scenario
• false propaganda, Applications
• FAQ bots, A Simple FAQ Bot-A Simple FAQ Bot, A Taxonomy of Chatbots
• Fast Healthcare Interoperability Resources (FHIR) standard, Outcome prediction and best practices, Medical Information Extraction and Analysis
• fastText classification models, An Example Scenario
• fastText embeddings (Facebook), Pre-trained word embeddings, Going Beyond Words, Subword Embeddings and fastText
• fastText library (Facebook), Subword Embeddings and fastText-Subword Embeddings and fastText
• fat-finger problem, Data Acquisition, Spelling Correction
• feature engineering, Advanced Processing-DL Pipeline, Building THE Model, Finding Better Features
• feature extraction, Feature Engineering, Text Representation
• (see also feature engineering)
• feature representation, Text Representation
• handcrafted, Handcrafted Feature Representations-Handcrafted Feature Representations
• feature selection, Finding Better Features-Finding Better Features
• features, Machine Learning, Deep Learning, and NLP: An Overview
• feedback, Components of a Search Engine, A Typical Enterprise Search Pipeline
• from customers (see review analysis)
• explicit, Case Study: Corporate Ticketing
• implicit, Case Study: Corporate Ticketing
• learning from, Case Study: Corporate Ticketing
• few-shot learning, Why Deep Learning Is Not Yet the Silver Bullet for NLP
• FHIR (Fast Healthcare Interoperability Resources) standard, Outcome prediction and best practices, Medical Information Extraction and Analysis
• figurative language, Ambiguity
• Figure Eight, One Pipeline, Many Classifiers, Document Embeddings, Explaining Classifier Predictions with Lime, No Training Data
• filtering adult content, Applications
• filtering, collaborative, Recommender Systems for Textual Data
• financial industry, Healthcare, Finance, and Law
• Financial PhraseBank, Financial sentiment
• financial sentiment, Financial sentiment-Financial sentiment
• financial services, NLP in the Real World, Finance and Law-Finance and Law
• estimated ML benefits, Finance and Law
• NLP applications, NLP Applications in Finance-Accounting and auditing
• FinBERT (BERT for financial text), Financial sentiment
• Finnish, Working with Other Languages
• fixed responses, Response Generation
• Flask, An Example Scenario
• formalism, Unique Challenges
• forms: data extraction from, IE Applications
• Freebase, Approaches to RE
• French, Working with Other Languages

• GATE (General Architecture for Text Engineering), Heuristics-Based NLP
• gated recurrent units (GRUs), Long short-term memory
• Gehrmann, Sebastian, Fake News
• generative chatbots, open-ended, Open-Ended Generative Chatbots-Open-Ended Generative Chatbots
• generative classifier, Logistic Regression
• gensim library, Pre-trained word embeddings, SkipGram, Word Embeddings, Setting Up a Summarizer: An Example
• KPE with, Implementing KPE
• text summarization with, Setting Up a Summarizer: An Example
• tutorial on LDA, Training a Topic Model: An Example
• Global AI survey (McKinsey), Peeking over the Horizon
• GloVe , Pre-trained word embeddings, Word Embeddings, Deep Learning for Text Classification, Text Representation for SMTD
• Gmail, NLP in the Real World, Temporal Information Extraction
• Gnip, Trending Topics
• goal-oriented dialogs, Goal-Oriented Dialog, Datasets
• Goldberg, Yoav, Text Representation
• good, social, Peeking over the Horizon
• Google
• Captcha tests, No Training Data
• data acquisition, Data Acquisition
• dataset search system, One Pipeline, Many Classifiers
• search engine, NLP in the Real World, Search and Information Retrieval, Components of a Search Engine, A Typical Enterprise Search Pipeline, Question-Answering Systems
• Word2vec model, Word Embeddings-Pre-trained word embeddings, SkipGram, SkipGram, Distributed Representations Beyond Words and Characters, Universal Text Representations, Word Embeddings, Text Representation for SMTD, An Example Scenario
• Google AI, Outcome prediction and best practices, Outcome prediction and best practices, Minimizing Technical Debt, Other Aspects
• Google AI for Social Good, Peeking over the Horizon
• Google Analytics, Practical Advice
• Google APIs, Using Existing Text Classification APIs, Case Study: Corporate Ticketing, Using a Machine Translation API: An Example
• Google Assistant, NLP in the Real World, Chatbots
• Google Cloud
• AutoML, Automating Machine Learning, Google Cloud AutoML and other techniques
• Dialogflow, PizzaStop Chatbot-Testing our agent, Utilizing Existing Frameworks
• Natural Language, Building Your Model, Using Existing Text Classification APIs
• NLP, Case Study
• task queues, Deployment
• Google Cloud Storage, Revisiting the NLP Pipeline: Deploying NLP Software
• Google Docs, NLP in the Real World
• Google Home, NLP: A Primer
• Google Knowledge Graph, NLP in the Real World, Named Entity Disambiguation and Linking
• Google News, IE Applications
• Google Search, NLP Tasks, Search and Information Retrieval, Minimizing Technical Debt
• Google Translate, NLP in the Real World, NLP Tasks, Ambiguity, Topics in Brief, Machine Translation, Using a Machine Translation API: An Example
• language identification, Applications
• grammar, Unique Challenges
• grammar-correction tools, NLP in the Real World, Handcrafted Feature Representations
• Grammarly, NLP in the Real World, Handcrafted Feature Representations
• ground truth, Machine Learning, Deep Learning, and NLP: An Overview, Intrinsic Evaluation
• Grover (AllenNLP), Fake News
• GRUs (gated recurrent units), Long short-term memory

• handcrafted feature representations, Handcrafted Feature Representations-Handcrafted Feature Representations
• Harvard University, Fake News
• HARVEST system, HARVEST: Longitudinal report understanding-HARVEST: Longitudinal report understanding
• health assistants, Health Assistants-Health Assistants
• health outcomes, Outcome prediction and best practices
• health records, Health and Medical Records
• healthcare, NLP in the Real World, Healthcare-Medical Information Extraction and Analysis
• chatbots in, Applications
• Google APIs, Case Study: Corporate Ticketing
• NLP applications, Healthcare
• patient prioritization and billing, Patient Prioritization and Billing
• question answering for, Question answering for health-Question answering for health
• HealthTermFinder, HARVEST: Longitudinal report understanding
• Heinzerling, Benjamin, Other Aspects
• heuristics, Approaches to NLP-Heuristics-Based NLP, Going Beyond Words
• combining with ML, Building Your Model
• reapplying, Building THE Model
• simple, Start with Simple Heuristics
• Hidden Markov Model (HMM), Hidden Markov Model-Hidden Markov Model
• Hindi, Working with Other Languages
• HMM (Hidden Markov Model), Hidden Markov Model-Hidden Markov Model
• Horowitz, Andreessen, E-Commerce and Retail
• HTML parsing and cleanup, HTML Parsing and Cleanup-HTML Parsing and Cleanup
• Huang, Jensen, Healthcare, Finance, and Law
• Hubbard, Elbert Green, Chatbots
• human-in-the-loop, Human-in-the-Loop-Human-in-the-Loop

• i2b2 datasets, Question answering for health
• IBM Research, Handcrafted Feature Representations, Named Entity Disambiguation and Linking, Question answering for health
• IBM Watson, NLP in the Real World, Building Your Model, Named Entity Disambiguation and Linking, Looking for Deeper Answers
• RE with, RE with the Watson API-RE with the Watson API
• ICD-10-CM, Medical Information Extraction and Analysis
• IDF (inverse document frequency), TF-IDF
• IE (see information extraction)
• image match, Image match
• image representation, Text Representation
• images: text extraction from, System-Specific Error Correction
• Imbalanced-Learn, Naive Bayes Classifier
• IMDB dataset, Understanding Twitter Sentiment
• in-memory databases, Universal Text Representations
• indexers, Components of a Search Engine
• indexing, Search and Information Retrieval, Components of a Search Engine, A Typical Enterprise Search Pipeline, Setting Up a Search Engine: An Example
• indirect attribute extraction, Indirect attribute extraction-Indirect attribute extraction
• Infermedica API, Health Assistants, Health Assistants
• information, Information Extraction
• information extraction (IE), NLP Tasks, Information Extraction-Practical Advice
• advanced tasks, Other Advanced IE Tasks-Template Filling
• case study, Case Study-Case Study
• general pipeline, The General Pipeline for IE-The General Pipeline for IE
• historical background, Information Extraction
• legal, Legal entity extraction with LexNLP-Legal entity extraction with LexNLP
• from loan agreements, Risk assessments
• medical, Medical Information Extraction and Analysis-Medical Information Extraction and Analysis
• open, Approaches to RE, Approaches to RE
• real-world applications, IE Applications-IE Applications
• tasks, IE Tasks-IE Tasks
• temporal, IE Tasks, Temporal Information Extraction-Temporal Information Extraction
• typical pipeline, Keyphrase Extraction
• information retrieval, NLP Tasks, Search and Information Retrieval-A Case Study: Book Store Search
• Instagram, Social Media
• intelligent machines, Peeking over the Horizon
• intelligent tutoring systems, NLP in the Real World
• interactive learning, Rasa NLU
• interpretability, Outcome prediction and best practices, Troubleshooting and Interpretability
• inverted indexes, Components of a Search Engine
• iNYP system, HARVEST: Longitudinal report understanding-HARVEST: Longitudinal report understanding

• Japanese, Working with Other Languages
• JAPE (Java Annotation Patterns Engine), Heuristics-Based NLP
• Jeopardy!, NLP in the Real World, Looking for Deeper Answers
• Jordan, Jeff, E-Commerce and Retail

• Kaggle, Understanding Twitter Sentiment
• Karpathy, Andrej, Other Aspects
• KDD (Knowledge Discovery and Data Mining) process, The KDD Process-The KDD Process
• key performance indicators (KPIs), A Pipeline for Building Text Classification Systems, Iterating Existing Models, Monitoring
• key terms, Word Cloud
• key words, Topic Modeling
• (see also topics)
• keyphrase extraction (KPE), IE Tasks, Keyphrase Extraction-Practical Advice, Practical Advice
• keyword extraction, IE Tasks
• knowledge bases, NLP in the Real World
• Knowledge Discovery and Data Mining (KDD) process, The KDD Process-The KDD Process
• knowledge-based question answering, Looking for Deeper Answers
• Koffka, Kurt, NLP Pipeline
• Korean, Working with Other Languages
• KPE (see keyphase extraction)
• KPIs (key performance indicators), A Pipeline for Building Text Classification Systems, Iterating Existing Models, Monitoring
• ktrain, Text Classification with Large, Pre-Trained Language Models
• Kubernetes, An Example Scenario

• labeled data, Data Acquisition
• labeling, Data and Timing
• labels, Machine Learning, Deep Learning, and NLP: An Overview, Intrinsic Evaluation
• language
• ambiguity in, Ambiguity-Ambiguity
• building blocks of, What Is Language?-Context
• characteristics that make NLP challenging, Why Is NLP Challenging?-Diversity across languages
• diversity across, Diversity across languages
• figurative, Ambiguity
• non-English languages, Language detection, Working with Other Languages-Working with Other Languages
• language detection, Language detection
• language identification, Applications
• language learning apps, NLP in the Real World
• language modeling, NLP Tasks, Universal Text Representations
• language models, large, pre-trained, Text Classification with Large, Pre-Trained Language Models-Text Classification with Large, Pre-Trained Language Models
• LARA (latent rating regression analysis), Connecting Overall Ratings to Aspects
• large, pre-trained language models, Text Classification with Large, Pre-Trained Language Models-Text Classification with Large, Pre-Trained Language Models
• latent attribute extraction, Latent attribute extraction from reviews
• latent Dirichlet allocation (LDA), Topic Modeling-Topic Modeling, Unsupervised approach, Latent attribute extraction from reviews
• latent rating regression analysis (LARA), Connecting Overall Ratings to Aspects
• latent semantic analysis (LSA), Topic Modeling
• Latin, Working with Other Languages
• law (see legal services)
• LDA (latent Dirichlet allocation), Topic Modeling-Topic Modeling, Unsupervised approach, Latent attribute extraction from reviews
• leadership, Team
• learning
• interactive, Rasa NLU
• with no or less data, Learning with No or Less Data and Adapting to New Domains-Less Training Data: Active Learning and Domain Adaptation
• learning and assessment tools, NLP in the Real World
• legal discovery, NLP and the Legal Landscape
• legal entity extraction, Legal entity extraction with LexNLP-Legal entity extraction with LexNLP
• legal research, NLP and the Legal Landscape
• legal services, Healthcare, Finance, and Law, Finance and Law, NLP and the Legal Landscape-Legal entity extraction with LexNLP
• chatbots, Applications
• lemmatization, Stemming and lemmatization
• length of text, Unique Challenges
• lexemes, Morphemes and lexemes
• lexicon-based sentiment analysis, A Simple Classifier Without the Text Classification Pipeline
• LexNLP, Legal entity extraction with LexNLP-Legal entity extraction with LexNLP
• LexRank, Understanding Aspects
• Lime, Interpreting Text Classification Models, Troubleshooting and Interpretability, Troubleshooting and Interpretability
• linguistics, What Is Language?-Context
• link prediction, Product linking
• LinkedIn, Identifying Memes
• Lipton, Zachary, Why Deep Learning Is Not Yet the Silver Bullet for NLP
• loan agreements: entity extraction from, Risk assessments
• locational features, Direct attribute extraction
• logistic regression, Logistic Regression
• long short-term memory networks (LSTMs), Long short-term memory, Deep Learning for Text Classification, LSTMs for Text Classification-LSTMs for Text Classification, Fake News
• LSTM-CRF, Direct attribute extraction
• lowercasing, Frequent Steps
• LSA (latent semantic analysis), Topic Modeling
• LSI (Latent Semantic Indexing), Case Study
• LSTM-CRF, Direct attribute extraction
• LSTMs (long short-term memory networks), Long short-term memory, Deep Learning for Text Classification, LSTMs for Text Classification-LSTMs for Text Classification, Fake News
• Lucidworks, Product Categorization and Taxonomy

• machine intelligence, Machine Learning, Deep Learning, and NLP: An Overview
• (see also artificial intelligence)
• machine learning (ML), The End-to-End NLP Process
• automated, Automating Machine Learning-Google Cloud AutoML and other techniques
• classical, Classical NLP/ML Pipeline
• estimated benefits, Finance and Law
• feature engineering, Classical NLP/ML Pipeline
• future directions, Peeking over the Horizon-Peeking over the Horizon
• landmark work, Minimizing Technical Debt
• for NLP, Machine Learning for NLP-Conditional random fields
• overview, Machine Learning, Deep Learning, and NLP: An Overview-Machine Learning, Deep Learning, and NLP: An Overview
• machine translation (MT), NLP Tasks, Topics in Brief, Machine Translation-Practical Advice
• with APIs, Using a Machine Translation API: An Example-Using a Machine Translation API: An Example
• practical advice, Practical Advice-Practical Advice
• use cases, Machine Translation
• machine translation services, NLP in the Real World
• Malayalam, Working with Other Languages
• MAP (Mean Average Precision), Intrinsic Evaluation, Intrinsic Evaluation
• MAPE (Mean Absolute Percentage Error), Intrinsic Evaluation
• markup elements: removing, Removing markup elements
• McAuley, Julian, A Case Study: Substitutes and Complements
• McKinsey & Company, Peeking over the Horizon
• Mechanical Turk, Datasets, Attribute Extraction
• medical care (see healthcare)
• medical information extraction and analysis, Medical Information Extraction and Analysis-Medical Information Extraction and Analysis
• medical records, Health and Medical Records
• Meena system, Other Aspects
• memes, Identifying Memes-Identifying Memes
• mental healthcare monitoring, Applications, Mental Healthcare Monitoring-Mental Healthcare Monitoring
• mental health–related issues, Mental Healthcare Monitoring
• Message Understanding Conferences (US Navy), Information Extraction
• metadata, A Case Study: Book Store Search, Data and Timing
• METEOR, Intrinsic Evaluation
• metrics, A Good Process
• Microsoft, Advanced Processing
• AI for social good, Peeking over the Horizon
• anomaly detection, Monitoring
• data acquisition, Data Acquisition
• reference pipelines for MLOps, Revisiting the NLP Pipeline: Deploying NLP Software
• sentiment analysis APIs, Using Existing Text Classification APIs
• translation API, Using a Machine Translation API: An Example
• Microsoft Azure, Microsoft Team Data Science Process
• AutoML, Google Cloud AutoML and other techniques
• Blob Storage, Revisiting the NLP Pipeline: Deploying NLP Software
• Cognitive Services, Building Your Model, An Example Scenario
• Machine Learning, Google Cloud AutoML and other techniques
• NEL with, Named Entity Disambiguation and Linking-NEL Using Azure API
• Text Analytics API, NEL Using Azure API-NEL Using Azure API
• Microsoft Cortana, NLP in the Real World
• Microsoft Outlook, NLP in the Real World
• Microsoft Research, An Example Scenario
• Microsoft REST API, Spelling Correction
• Microsoft Team Data Science Process (TDSP), The Data Science Process, Microsoft Team Data Science Process-Microsoft Team Data Science Process
• Microsoft Word, NLP in the Real World, Handcrafted Feature Representations
• Milne, A.A., Text Classification
• Minimum Viable Product (MVP) approach, The Philosophy
• misinformation, Fake News
• (see also fake news)
• MIT, Question answering for health
• MIT Sloan, Peeking over the Horizon
• MITIE, Building an NER System
• ML (see machine learning)
• MLOps, Revisiting the NLP Pipeline: Deploying NLP Software
• model compression, An Example Scenario-An Example Scenario
• model ensembling, Building THE Model, Building THE Model
• model packaging, Revisiting the NLP Pipeline: Deploying NLP Software
• model scaling, Revisiting the NLP Pipeline: Deploying NLP Software
• model serving, Revisiting the NLP Pipeline: Deploying NLP Software
• model size, An Example Scenario
• model stacking, Building THE Model, Building THE Model
• modeling, Modeling-Building THE Model, An Example Scenario
• with APIs, Building Your Model
• approaches to, Building THE Model-Building THE Model
• combining heuristics with, Building Your Model
• customization, Rasa NLU
• deployment, Deployment
• evaluation of, Peeking over the Horizon
• goodness of fit, Evaluation
• interpretability, Outcome prediction and best practices
• iterating existing models, Iterating Existing Models
• monitoring, Monitoring, Monitoring-Monitoring
• recommendations for, Minimizing Technical Debt
• reproducibility, Code and Model Reproducibility
• sequence-to-sequence (seq2seq) models, End-to-End Approach
• slot identification with, Models-Models
• state-of-the-art (SOTA) models, A Good Process
• strategies for, Building THE Model
• testing models, Troubleshooting and Interpretability, Troubleshooting and Interpretability-Troubleshooting and Interpretability
• updating, Model Updating, Building and Maintaining a Mature System
• web service–based, Monitoring
• Molnar, Christoph, Troubleshooting and Interpretability
• monitoring, Monitoring, Monitoring-Monitoring
• Moore, Anthony, The End-to-End NLP Process
• morphemes, Morphemes and lexemes
• morphological analysis, Morphemes and lexemes, Working with Other Languages
• MRR (Mean Reciprocal Rank), Intrinsic Evaluation, Intrinsic Evaluation
• MT (see machine translation)
• multi-document summarization, Text Summarization
• multiclass classification, Text Classification
• multilingual writing, Unique Challenges
• MultiWoZ dataset, Datasets
• MVP approach (see Minimum Viable Product approach)

• n-grams, Bag of N-Grams
• Nadella, Satya, Advanced Processing
• Naive Bayes, Naive Bayes, Data and Timing
• Naive Bayes classifier, Naive Bayes Classifier-Naive Bayes Classifier
• named entity disambiguation (NED), Named Entity Disambiguation and Linking
• named entity disambiguation and linking, IE Tasks
• named entity linking (NEL), NEL Using Azure API-NEL Using Azure API
• named entity recognition (NER), An NLP Walkthrough: Conversational Agents, Advanced Processing, IE Tasks, Named Entity Recognition-Practical Advice
• with active learning, NER Using Active Learning-NER Using Active Learning
• building, Building an NER System-Building an NER System
• examples, Named Entity Recognition
• with existing libraries, NER Using an Existing Library
• in finance, Risk assessments
• HealthTermFinder, HARVEST: Longitudinal report understanding
• practical advice, Practical Advice-Practical Advice
• rule-based, Building an NER System
• typical training data, Building an NER System
• Natty, Temporal Information Extraction
• natural language generation, A Pipeline for Building Dialog Systems
• Natural Language Processing (NLP), Preface
• approaches to, Approaches to NLP-Why Deep Learning Is Not Yet the Silver Bullet for NLP
• case studies, An NLP Walkthrough: Conversational Agents-An NLP Walkthrough: Conversational Agents, Case Study-Case Study
• classical, Feature Engineering, Classical NLP/ML Pipeline
• core applications, NLP in the Real World
• deployment, Revisiting the NLP Pipeline: Deploying NLP Software-Revisiting the NLP Pipeline: Deploying NLP Software
• DL-based, Feature Engineering
• end-to-end process, The End-to-End NLP Process-Other Aspects
• heuristics-based, Approaches to NLP-Heuristics-Based NLP
• key decision points, Revisiting the NLP Pipeline: Deploying NLP Software
• overview, NLP: A Primer-Wrapping Up
• pipeline (see pipeline)
• project life cycle, Other Aspects
• for social data, NLP for Social Data-Customer Support on Social Channels
• software applications, NLP in the Real World-NLP in the Real World
• tasks and applications, NLP: A Primer, NLP Tasks-NLP Tasks
• tasks organized according to relative difficulty, NLP Tasks
• Natural Language Tool Kit (NLTK)
• lemmatization with, Stemming and lemmatization
• POS tagging, Advanced Processing
• sentence segmentation with, Sentence segmentation
• stemming with, Stemming and lemmatization
• stop words, Frequent Steps
• tokenizer, Tokenizer for SMTD
• word tokenization with, Word tokenization-Word tokenization
• natural language understanding (NLU), A Pipeline for Building Dialog Systems, Deep Dive into Components of a Dialog System
• case study, An NLP Walkthrough: Conversational Agents
• Rasa, Rasa NLU-Rasa NLU
• Navajo, Practical Advice
• NCRF++, Building an NER System
• NED (named entity disambiguation), Named Entity Disambiguation and Linking
• NEL (named entity linking), NEL Using Azure API-NEL Using Azure API
• NER (see named entity recognition)
• Netflix, Data Acquisition, Topics in Brief, Recommender Systems for Textual Data
• neural embeddings, Using Neural Embeddings in Text Classification-Document Embeddings
• neural text representation, Universal Text Representations
• New York Presbyterian Hospital, HARVEST: Longitudinal report understanding
• iNYP, HARVEST: Longitudinal report understanding-HARVEST: Longitudinal report understanding
• news and content discovery, Applications
• news classification, Applications, Fake News
• news: tagging, IE Applications
• NIST Text Analysis Conference, Information Extraction
• NLP (see Natural Language Processing)
• NLPAug, Data Acquisition
• NLTK (see Natural Language Tool Kit)
• nltk.tokenize.TweetTokenizer, Tokenizer for SMTD
• NLU (see natural language understanding)
• noise, Customer Support on Social Channels
• adding to data, Data Acquisition
• examples, Customer Support on Social Channels
• on social media, Unique Challenges
• non-English language detection, Language detection
• non-English language processing, Working with Other Languages-Working with Other Languages
• non-text data, Handling non-text data
• normalization
• temporal IE and normalization, Temporal Information Extraction
• text, Text normalization, A Typical Enterprise Search Pipeline
• Unicode, Unicode Normalization
• Norvig, Peter, Minimizing Technical Debt
• numbers: removing, Frequent Steps

• Occam’s razor, Why Deep Learning Is Not Yet the Silver Bullet for NLP
• OCR (optical character recognition), System-Specific Error Correction, IE Applications
• Olah, Christopher, Long short-term memory
• one-hot encoding, One-Hot Encoding-One-Hot Encoding
• online support, Applications
• ONNX, Revisiting the NLP Pipeline: Deploying NLP Software
• OntoNotes, Building an NER System
• OOV (out of vocabulary) problem, One-Hot Encoding, Going Beyond Words, Subword Embeddings and fastText, Unique Challenges-Unique Challenges, Text Representation for SMTD
• Oozie (Apache), Revisiting the NLP Pipeline: Deploying NLP Software
• open formats, Revisiting the NLP Pipeline: Deploying NLP Software
• open IE, Approaches to RE, Approaches to RE
• Open Mind Common Sense, Heuristics-Based NLP
• open-ended generative chatbots, Open-Ended Generative Chatbots-Open-Ended Generative Chatbots
• opinion mining, Applications
• optical character recognition (OCR), System-Specific Error Correction, IE Applications
• out of vocabulary (OOV) problem, One-Hot Encoding, Going Beyond Words, Subword Embeddings and fastText, Unique Challenges-Unique Challenges, Text Representation for SMTD
• outcome prediction, Outcome prediction and best practices-Outcome prediction and best practices

• packaging, Revisiting the NLP Pipeline: Deploying NLP Software
• parse trees, Syntax
• Parsedatetime, Temporal Information Extraction
• Parsey McParseface Tagger, Advanced Processing
• part-of-speech (POS) tagging, Hidden Markov Model, Advanced Processing, Advanced Processing
• Partnership on AI, Other Aspects
• PDF documents: text extraction from, System-Specific Error Correction
• PDFMiner, System-Specific Error Correction
• Penn Treebank dataset, Automating Machine Learning
• perplexity, Intrinsic Evaluation
• personnel, Team-Team
• pharmacovigilance, Pharmacovigilance
• phonemes, What Is Language?, Phonemes
• phonetic typing, Unique Challenges
• Pineau, Joelle, Code and Model Reproducibility
• pipeline, NLP Pipeline-Wrapping Up
• generic components, NLP Pipeline
• key stages, NLP Pipeline-NLP Pipeline
• plagiarism detection, NLP in the Real World
• PLSA (probabilistic latent semantic analysis), Topic Modeling
• polarity, Understanding Twitter Sentiment
• Polyglot Python Library, Language detection
• Porter Stemmer, Stemming and lemmatization
• POS (part-of-speech) tagging, Hidden Markov Model, Advanced Processing, Advanced Processing
• post-modeling phases, Post-Modeling Phases-Model Updating
• pragmatics, Context
• pre-processing, System-Specific Error Correction-Advanced Processing, Building Your Model, Mental Healthcare Monitoring, The KDD Process
• advanced processing, Advanced Processing
• common steps, Stemming and lemmatization
• frequent steps, Frequent Steps-Stemming and lemmatization
• other steps, Other Pre-Processing Steps-Code mixing and transliteration
• SMTD, Pre-Processing SMTD-Nonstandard spellings
• pre-trained language models, large, Text Classification with Large, Pre-Trained Language Models-Text Classification with Large, Pre-Trained Language Models
• pre-training, Transformers, Universal Text Representations
• precision, Intrinsic Evaluation
• predicting health outcomes, Outcome prediction and best practices-Outcome prediction and best practices
• pregex, Heuristics-Based NLP
• preliminaries, Preliminaries-Word tokenization
• probabilistic latent semantic analysis (PLSA), Topic Modeling
• Prodigy, Less Training Data: Active Learning and Domain Adaptation, Case Study: Corporate Ticketing, NER Using Active Learning
• product catalogs (see catalogs)
• product categorization and taxonomy, Product Categorization and Taxonomy-Product Categorization and Taxonomy
• product deduplication and matching, Product Deduplication and Matching-Image match
• product enrichment, Product Enrichment-Product Enrichment
• product intervention, Data Acquisition
• product linking, Product linking
• product recommendations, Product Recommendations, Recommendations for E-Commerce
• (see also recommendations)
• product search, Product Search
• propaganda, false, Applications
• public datasets, Data Acquisition, Case Study: Corporate Ticketing, Understanding Twitter Sentiment
• PubMed, Medical Information Extraction and Analysis
• punctuation: removing, Frequent Steps
• PwC, Accounting and auditing
• PyPDF, System-Specific Error Correction
• Python, System-Specific Error Correction

• quality of data, Data and Timing
• quantity of data, Data and Timing
• query-focused summarization, Text Summarization
• question answering (QA), NLP Tasks, Topics in Brief
• with deep neural networks, Looking for Deeper Answers
• knowledge-based, Looking for Deeper Answers
• question-answering (QA) systems, Question-Answering Systems-Looking for Deeper Answers
• dataset creation framework , Question answering for health
• dataset generation with existing annotations, Question answering for health
• developing, Developing a Custom Question-Answering System
• DL-based, Looking for Deeper Answers
• for health, Question answering for health-Question answering for health

• random forest, Data and Timing
• ranking, Intrinsic Evaluation, Components of a Search Engine, A Typical Enterprise Search Pipeline
• Rao, Delip, Other Aspects
• Rasa, Building an NER System, Rasa NLU-Rasa NLU, Utilizing Existing Frameworks
• Rasa NLP, An Example Scenario
• Rasa NLU, Rasa NLU-Rasa NLU
• recall, Intrinsic Evaluation
• Recall at rank K, Intrinsic Evaluation
• Recall-Oriented Understudy for Gisting Evaluation (ROUGE), Intrinsic Evaluation, Practical Advice
• receipts: data extraction from, IE Applications
• recommendation engines, Recommendations for E-Commerce
• (see also recommender systems)
• recommendations, Topics in Brief
• case study, A Case Study: Substitutes and Complements-Product linking
• for e-commerce, Recommendations for E-Commerce-Product linking
• product, Product Recommendations
• techniques for, Recommendations for E-Commerce
• recommender systems, Recommender Systems for Textual Data-Practical Advice, A Case Study: Substitutes and Complements
• creating, Creating a Book Recommender System: An Example-Creating a Book Recommender System: An Example
• e-commerce, Product Recommendations, Recommendations for E-Commerce
• examples, Creating a Book Recommender System: An Example-Creating a Book Recommender System: An Example
• practical advice, Practical Advice-Practical Advice
• recurrent neural networks (RNNs), Recurrent neural networks, Universal Text Representations, Deep Learning for Text Classification, Direct attribute extraction
• Reddit autotldr bot, Summarization Use Cases
• Redis database, Universal Text Representations
• RegexNER (Stanford NLP), Building an NER System, NER Using Active Learning
• regression techniques, Machine Learning for NLP
• regression, logistic, Logistic Regression
• regular expressions (regex), Heuristics-Based NLP
• reinforcement learning, Machine Learning, Deep Learning, and NLP: An Overview
• deep, Deep Reinforcement Learning for Dialogue Generation-Human-in-the-Loop
• for dialogue generation, Deep Reinforcement Learning for Dialogue Generation-Human-in-the-Loop
• related queries, Search and Information Retrieval
• relation extraction (see relationship extraction)
• relationship extraction (RE), Advanced Processing, IE Tasks, Relationship Extraction-RE with the Watson API
• approaches to, Approaches to RE-Approaches to RE
• example, Relationship Extraction
• unsupervised, Approaches to RE
• with Watson API, RE with the Watson API-RE with the Watson API
• replacing entities, Data Acquisition
• representative data, Data and Timing
• reproducibility, Code and Model Reproducibility
• research, legal, NLP and the Legal Landscape
• response generation, Response Generation
• automatic, Response Generation
• case study, An NLP Walkthrough: Conversational Agents
• fixed responses, Response Generation
• templates-based, Response Generation
• retail (see e-commerce and retail)
• Reuters, Financial sentiment
• review analysis, Review Analysis, Review Analysis-Understanding Aspects
• connecting overall ratings to aspects, Connecting Overall Ratings to Aspects-Connecting Overall Ratings to Aspects
• example pipeline, Understanding Aspects
• latent attribute extraction from, Latent attribute extraction from reviews
• substitutes and complements based on, A Case Study: Substitutes and Complements
• risk assessments, Risk assessments-Risk assessments
• Ritter, Allen, Text Representation for SMTD
• RMSE (Root Mean Squared Error), Intrinsic Evaluation
• RNNs (recurrent neural networks), Recurrent neural networks, Universal Text Representations, Deep Learning for Text Classification, Direct attribute extraction
• ROI (return on investment), Other Aspects
• Rosette Text Analytics, Relationship Extraction
• ROSS Intelligence service, NLP and the Legal Landscape
• ROUGE (Recall-Oriented Understudy for Gisting Evaluation), Intrinsic Evaluation, Practical Advice
• Ruder, Sebastian, SkipGram
• rule-based systems, Heuristics-Based NLP
• rules, Heuristics-Based NLP
• rumor/fake news detection, Applications
• RxNorm, Medical Information Extraction and Analysis

• SAFE (simple agreement for future equity), Legal entity extraction with LexNLP
• SageMaker (AWS), Revisiting the NLP Pipeline: Deploying NLP Software, An Example Scenario
• Sanders, Niek, Understanding Twitter Sentiment
• sarcasm, Universal Text Representations
• scaling, Revisiting the NLP Pipeline: Deploying NLP Software
• scanned documents: text extraction from, System-Specific Error Correction
• scraping data, Data Acquisition
• Scrapy, HTML Parsing and Cleanup, Components of a Search Engine
• search, Topics in Brief, Search and Information Retrieval-A Case Study: Book Store Search
• faceted, Search in E-Commerce-Search in E-Commerce
• focused, Product Search
• product, Product Search
• search engines, NLP in the Real World
• case study, A Case Study: Book Store Search-A Case Study: Book Store Search
• components, Components of a Search Engine-Components of a Search Engine
• in e-commerce, Search in E-Commerce-Search in E-Commerce
• enterprise, Search and Information Retrieval
• example build, Setting Up a Search Engine: An Example-Setting Up a Search Engine: An Example
• features that use NLP, Search and Information Retrieval
• feedback, A Typical Enterprise Search Pipeline
• generic, Search and Information Retrieval
• managed services, A Case Study: Book Store Search
• query processing and execution, A Typical Enterprise Search Pipeline
• ranking, A Typical Enterprise Search Pipeline
• search results classification, Search and Information Retrieval
• searchers, Components of a Search Engine
• seed words or seed lexicons, Aspect-Level Sentiment Analysis
• self-attention, Transformers, Transformers
• semantics, Context, Text Representation
• Semantics3, Product Categorization and Taxonomy
• semi-supervised learning, Machine Learning, Deep Learning, and NLP: An Overview
• sentence segmentation, Sentence segmentation
• sentiment analysis, An NLP Walkthrough: Conversational Agents, Applications, Sentiment Analysis-Sentiment Analysis
• aspect-based, Applications-Applications
• aspect-level, Aspect-Level Sentiment Analysis-Unsupervised approach
• financial, Financial sentiment-Financial sentiment
• lexicon-based, A Simple Classifier Without the Text Classification Pipeline
• with social media data, Applications, Understanding Twitter Sentiment-Understanding Twitter Sentiment
• supervised approach, Supervised approach
• testing, Troubleshooting and Interpretability
• tracking changes over time, Understanding Twitter Sentiment
• unsupervised , Unsupervised approach-Unsupervised approach
• sentiment analysis APIs, Using Existing Text Classification APIs
• Sentiment Analysis: Emotion in Text dataset (Figure Eight), Document Embeddings
• sequence classification, Building an NER System
• sequence labeling, Building an NER System
• sequence-to-sequence models (seq2seq) , End-to-End Approach
• Shakespeare, William, Information Extraction
• Shap, Troubleshooting and Interpretability
• Shenfeld, Daniel, Other Aspects
• shopping (see e-commerce and retail)
• short-term memory, long, Long short-term memory
• shorthand, Spelling Correction
• Siamese networks, Title match
• simple agreement for future equity (SAFE), Legal entity extraction with LexNLP
• single-document summarization, Text Summarization
• Singlish, Code mixing and transliteration
• siren.io platform, NLP and the Legal Landscape
• SkipGram, SkipGram-SkipGram
• slots or entities, Dialog Systems in Detail, Identifying Slots-Identifying Slots, Slot identification-Models
• small datasets, Why Deep Learning Is Not Yet the Silver Bullet for NLP
• SMTD (see social media text data)
• snippet extraction, Search and Information Retrieval
• SNIPS dataset, Datasets, Loading the dataset
• Snips platform, Loading the dataset
• Snorkel software, Data Acquisition, No Training Data
• Snow, C.P., Peeking over the Horizon
• social datasets, Mental Healthcare Monitoring
• social good, Peeking over the Horizon
• social media, NLP in the Real World, Applications, Social Media-Wrapping Up
• data generated in, Social Media
• example posts, Unique Challenges, Unique Challenges
• information extraction from, IE Applications
• social media text data (SMTD), Applications
• applications that use, Applications-Applications
• NLP for, NLP for Social Data-Customer Support on Social Channels
• normalization of, Text normalization
• pre-processing, Pre-Processing SMTD-Nonstandard spellings
• sentiment analysis with, Understanding Twitter Sentiment-Understanding Twitter Sentiment
• special characters, Unique Challenges
• text representation for, Text Representation for SMTD-Text Representation for SMTD
• tokenizers for, Tokenizer for SMTD
• unique challenges, Unique Challenges-Unique Challenges
• software applications, NLP in the Real World-NLP in the Real World
• software deployment, Revisiting the NLP Pipeline: Deploying NLP Software-Revisiting the NLP Pipeline: Deploying NLP Software
• SOTA (state-of-the-art) models, A Good Process
• spaCy library, NER Using an Existing Library
• EntityRuler, Building an NER System, NER Using Active Learning
• lemmatization with, Stemming and lemmatization
• NER with, NER Using an Existing Library
• POS tagging, Advanced Processing
• pre-processing with, Advanced Processing
• rule-based matcher, Start with Simple Heuristics, Building Your Model
• source code, Text normalization
• tokenization with, Word tokenization
• Spanish, Working with Other Languages
• special characters, Unique Challenges
• speech recognition, An NLP Walkthrough: Conversational Agents, A Pipeline for Building Dialog Systems
• speech representation, Text Representation
• speech synthesis, An NLP Walkthrough: Conversational Agents, A Pipeline for Building Dialog Systems
• spell checking (see spelling correction)
• spelling correction, Spelling Correction-Spelling Correction, Search and Information Retrieval
• spelling, nonstandard, Working with Other Languages, Unique Challenges, Nonstandard spellings-Nonstandard spellings
• spelling- and grammar-correction tools, NLP in the Real World, Handcrafted Feature Representations
• spelling-correction libraries, Nonstandard spellings
• split-joined words, Split-joined words
• Spoke corporate ticketing system (case study), Case Study: Corporate Ticketing-Case Study: Corporate Ticketing
• SpotDraft, NLP and the Legal Landscape
• SQuAD dataset, Looking for Deeper Answers
• Stack Overflow web pages: text extraction from, HTML Parsing and Cleanup
• stacking, Building THE Model, Building THE Model
• Stanford Medicine, Outcome prediction and best practices
• Stanford Named Entity Recognizer (NER), NER Using an Existing Library
• Stanford Natural Language Processing Group
• CoreNLP output, Advanced Processing
• GloVe, Pre-trained word embeddings, Text Representation for SMTD
• RegexNER, Building an NER System, NER Using Active Learning
• SUTime, Temporal Information Extraction
• TokensRegex, Heuristics-Based NLP, Start with Simple Heuristics
• state-of-the-art (SOTA) models, A Good Process
• Steinhardt, Jacob, Why Deep Learning Is Not Yet the Silver Bullet for NLP
• stemming, Stemming and lemmatization
• stop words, Frequent Steps, Topic Modeling
• storage, Revisiting the NLP Pipeline: Deploying NLP Software
• subjectivity, Understanding Twitter Sentiment
• substitutes, A Case Study: Substitutes and Complements, A Case Study: Substitutes and Complements
• subword embeddings, Subword Embeddings and fastText-Subword Embeddings and fastText
• suicide, Mental Healthcare Monitoring
• summarization (see text summarization)
• summarization algorithms, Understanding Aspects
• Sumo Logic, Monitoring
• Sumy library, Setting Up a Summarizer: An Example
• supervised learning, Machine Learning, Deep Learning, and NLP: An Overview
• supervised machine learning techniques, Machine Learning for NLP
• support vector machines (SVMs), Support vector machine, Support vector machine, Support Vector Machine-Support Vector Machine
• SUTime (Stanford NLP), Temporal Information Extraction
• SVMs (support vector machines), Support vector machine, Support vector machine, Support Vector Machine-Support Vector Machine
• Swahili, Working with Other Languages
• Swiftype, A Case Study: Book Store Search
• synonym replacement, Data Acquisition
• Synsets in Wordnet, Data Acquisition
• syntax, Syntax
• synthetic data generation, Why Deep Learning Is Not Yet the Silver Bullet for NLP

• t-distributed Stochastic Neighboring Embedding (t-SNE), Visualizing Embeddings-Visualizing Embeddings
• TaggedDocument class, Document Embeddings
• target datasets, The KDD Process
• task managers, A Pipeline for Building Dialog Systems
• taxonomy, product, Product Categorization and Taxonomy-Product Categorization and Taxonomy
• Team Data Science Process (TDSP), The Data Science Process, Microsoft Team Data Science Process-Microsoft Team Data Science Process
• team personnel, Team-Team
• technical debt minimization, Minimizing Technical Debt-Minimizing Technical Debt
• template filling, IE Tasks, Template Filling-Template Filling
• templates-based response generation, Response Generation
• temporal IE and normalization, Temporal Information Extraction
• temporal information extraction, IE Tasks, Temporal Information Extraction-Temporal Information Extraction
• TensorBoard, Visualizing Embeddings, Visualizing Embeddings
• TensorFlow Extended, Troubleshooting and Interpretability
• TensorFlow model analysis (TFMA), Troubleshooting and Interpretability, Troubleshooting and Interpretability
• Term Frequency–Inverse Document Frequency (see TF-IDF)
• term vector models, Vector Space Models
• (see also vector space models)
• Tesseract, System-Specific Error Correction
• testing, Troubleshooting and Interpretability, Troubleshooting and Interpretability-Troubleshooting and Interpretability
• Text Analysis Conference (NIST), Information Extraction
• text categorization, Text Classification
• (see also text classification)
• text classification, NLP Tasks, Text Classification-Practical Advice, A Typical Enterprise Search Pipeline
• applications, Applications-Applications
• AWS pipeline, Revisiting the NLP Pipeline: Deploying NLP Software
• case study, Case Study: Corporate Ticketing-Case Study: Corporate Ticketing
• CNNs for, CNNs for Text Classification-CNNs for Text Classification
• definition, Text Classification
• DL for, Deep Learning for Text Classification-Text Classification with Large, Pre-Trained Language Models
• with document embeddings, Document Embeddings-Document Embeddings
• example scenario, An Example Scenario-An Example Scenario
• with existing APIs or libraries, Using Existing Text Classification APIs, Case Study: Corporate Ticketing
• interpreting models, Interpreting Text Classification Models-Explaining Classifier Predictions with Lime
• KPIs for, Monitoring
• with large, pre-trained language models, Text Classification with Large, Pre-Trained Language Models-Text Classification with Large, Pre-Trained Language Models
• logistic regression, Logistic Regression
• LSTMs for, LSTMs for Text Classification-LSTMs for Text Classification
• Naive Bayes classifier, Naive Bayes Classifier-Naive Bayes Classifier
• with neural embeddings, Using Neural Embeddings in Text Classification-Document Embeddings
• with no or less data, Case Study: Corporate Ticketing
• pipeline for building systems, A Pipeline for Building Text Classification Systems-Using Existing Text Classification APIs
• practical advice, Practical Advice-Practical Advice
• with public datasets, One Pipeline, Many Classifiers-One Pipeline, Many Classifiers, Case Study: Corporate Ticketing
• reasons for poor performance, Naive Bayes Classifier-Naive Bayes Classifier
• simple, A Simple Classifier Without the Text Classification Pipeline-A Simple Classifier Without the Text Classification Pipeline
• with SVMs, Support Vector Machine-Support Vector Machine
• text data, Unique Challenges
• text encoding, Unicode Normalization
• text extraction and cleanup, Text Extraction and Cleanup-System-Specific Error Correction
• text normalization, Text normalization, A Typical Enterprise Search Pipeline
• text recommendations (see recommendations)
• text representation, Feature Engineering, Text Representation-Wrapping Up
• basic vectorization approaches, Basic Vectorization Approaches-TF-IDF
• distributed representations, Distributed Representations-Distributed Representations Beyond Words and Characters
• distributional, Distributed Representations
• neural, Universal Text Representations
• for SMTD, Text Representation for SMTD-Text Representation for SMTD
• universal representations, Universal Text Representations-Universal Text Representations
• text summarization, NLP Tasks, Topics in Brief, Text Summarization-Practical Advice
• abstractive, Text Summarization, Practical Advice
• example setup, Setting Up a Summarizer: An Example-Setting Up a Summarizer: An Example
• extractive, Text Summarization, Practical Advice
• multi-document, Text Summarization
• practical advice, Practical Advice-Practical Advice
• query-focused, Text Summarization
• query-independent, Text Summarization
• single-document, Text Summarization
• use cases, Summarization Use Cases-Summarization Use Cases
• text-based chatbots, A Pipeline for Building Dialog Systems
• text-generation tasks, Intrinsic Evaluation-Intrinsic Evaluation
• textacy, Implementing KPE
• TextBlob toolkit, Understanding Twitter Sentiment, Nonstandard spellings
• TextEvaluator software (ETS), Handcrafted Feature Representations
• TextRank, Implementing KPE, Setting Up a Summarizer: An Example, Practical Advice
• textual data, Recommender Systems for Textual Data-Practical Advice
• TF (term frequency), TF-IDF
• TF-IDF (Term Frequency–Inverse Document Frequency), TF-IDF-TF-IDF, Search and Information Retrieval, Components of a Search Engine, A Typical Enterprise Search Pipeline, Topic Modeling, Text Representation for SMTD, HARVEST: Longitudinal report understanding
• TF-IDF–based word replacement, Data Acquisition
• TFMA (TensorFlow model analysis), Troubleshooting and Interpretability, Troubleshooting and Interpretability
• Thakur, Abhishek, Google Cloud AutoML and other techniques
• title match, Title match-Title match
• tokenization
• language-specific exceptions, Frequent Steps
• sentence (see sentence segmentation)
• for SMTD, Tokenizer for SMTD, Text Representation for SMTD
• specialized tokenizers, Tokenizer for SMTD
• tweet, Word tokenization, Document Embeddings
• word, Word tokenization-Word tokenization
• TokensRegex (Stanford NLP), Heuristics-Based NLP, Start with Simple Heuristics
• topic classification, Text Classification
• (see also text classification)
• topic detection, Text Classification
• trending topics, Applications, Trending Topics-Trending Topics
• topic hierarchy, Product linking
• topic modeling, NLP Tasks, Machine Learning, Deep Learning, and NLP: An Overview, Topics in Brief, Topic Modeling-What’s Next?, Unsupervised approach
• algorithms, Topic Modeling, Unsupervised approach
• example visualization, Topic Modeling
• training models, Training a Topic Model: An Example-Training a Topic Model: An Example
• use cases, What’s Next?-What’s Next?
• topic vectors, Latent attribute extraction from reviews
• topics, Topic Modeling
• training data, Machine Learning, Deep Learning, and NLP: An Overview
• crowdsourcing, No Training Data
• no data, No Training Data-No Training Data, Case Study: Corporate Ticketing
• transfer learning, Transformers, Building THE Model, Universal Text Representations, Less Training Data: Active Learning and Domain Adaptation-Less Training Data: Active Learning and Domain Adaptation
• transformers, Transformers-Transformers, Transformers
• translation (see machine translation)
• translation APIs, Using a Machine Translation API: An Example-Using a Machine Translation API: An Example
• translation cache, Practical Advice
• translation memory, Practical Advice
• transliteration, Code mixing and transliteration, Unique Challenges-Unique Challenges
• trending topic detection, Applications
• trending topics, Applications, Trending Topics-Trending Topics
• TripAdvisor, Product Search
• trolling, Memes and Fake News
• troubleshooting, Troubleshooting and Interpretability-Troubleshooting and Interpretability
• truly intelligent machines, Peeking over the Horizon
• Tsinghua University, An Example Scenario
• Turing, Alan, Chatbots
• Turkish, Working with Other Languages
• turnaround cycles, A Good Process
• Turnitin, NLP in the Real World
• Tweepy, Trending Topics-Trending Topics
• tweet tokenization, Word tokenization
• TweetTokenizer, Document Embeddings
• 20 Newsgroups dataset, Case Study: Corporate Ticketing
• Twikenizer, Tokenizer for SMTD
• Twitter, Data Acquisition, Document Embeddings, Social Media, Unique Challenges
• customer support on, Customer Support on Social Channels
• event extraction from, Event Extraction
• example posts, Unique Challenges
• rate limits, Trending Topics
• sentiment analysis with, Understanding Twitter Sentiment-Understanding Twitter Sentiment
• specialized tokenizers for data from, Tokenizer for SMTD
• Twitter Sentiment Corpus (Sanders), Understanding Twitter Sentiment
• twokenize, Tokenizer for SMTD, Text Representation for SMTD, Text Representation for SMTD
• Twokenizer, Tokenizer for SMTD