You are previewing Sentiment Analysis.
O'Reilly logo
Sentiment Analysis

Book Description

Sentiment analysis is the computational study of people's opinions, sentiments, emotions, and attitudes. This fascinating problem is increasingly important in business and society. It offers numerous research challenges but promises insight useful to anyone interested in opinion analysis and social media analysis. This book gives a comprehensive introduction to the topic from a primarily natural-language-processing point of view to help readers understand the underlying structure of the problem and the language constructs that are commonly used to express opinions and sentiments. It covers all core areas of sentiment analysis, includes many emerging themes, such as debate analysis, intention mining, and fake-opinion detection, and presents computational methods to analyze and summarize opinions. It will be a valuable resource for researchers and practitioners in natural language processing, computer science, management sciences, and the social sciences.

Table of Contents

  1. Cover
  2. Half title
  3. Title page
  4. Imprints page
  5. Contents
  6. Preface
  7. Acknowledgments
  8. 1 Introduction
    1. 1.1 Sentiment Analysis Applications
    2. 1.2 Sentiment Analysis Research
      1. 1.2.1 Different Levels of Analysis
      2. 1.2.2 Sentiment Lexicon and Its Issues
      3. 1.2.3 Analyzing Debates and Comments
      4. 1.2.4 Mining Intentions
      5. 1.2.5 Opinion Spam Detection and Quality of Reviews
    3. 1.3 Sentiment Analysis as Mini NLP
    4. 1.4 My Approach to Writing This Book
  9. 2 The Problem of Sentiment Analysis
    1. 2.1 Definition of Opinion
      1. 2.1.1 Opinion Definition
      2. 2.1.2 Sentiment Target
      3. 2.1.3 Sentiment of Opinion
      4. 2.1.4 Opinion Definition Simplified
      5. 2.1.5 Reason and Qualifier for Opinion
      6. 2.1.6 Objective and Tasks of Sentiment Analysis
    2. 2.2 Definition of Opinion Summary
    3. 2.3 Affect, Emotion, and Mood
      1. 2.3.1 Affect, Emotion, and Mood in Psychology
      2. 2.3.2 Affect, Emotion, and Mood in Sentiment Analysis
    4. 2.4 Different Types of Opinions
      1. 2.4.1 Regular and Comparative Opinions
      2. 2.4.2 Subjective and Fact-Implied Opinions
      3. 2.4.3 First-Person and Non-First-Person Opinions
      4. 2.4.4 Meta-Opinions
    5. 2.5 Author and Reader Standpoint
    6. 2.6 Summary
  10. 3 Document Sentiment Classification
    1. 3.1 Supervised Sentiment Classification
      1. 3.1.1 Classification Using Machine Learning Algorithms
      2. 3.1.2 Classification Using a Custom Score Function
    2. 3.2 Unsupervised Sentiment Classification
      1. 3.2.1 Classification Using Syntactic Patterns and Web Search
      2. 3.2.2 Classification Using Sentiment Lexicons
    3. 3.3 Sentiment Rating Prediction
    4. 3.4 Cross-Domain Sentiment Classification
    5. 3.5 Cross-Language Sentiment Classification
    6. 3.6 Emotion Classification of Documents
    7. 3.7 Summary
  11. 4 Sentence Subjectivity and Sentiment Classification
    1. 4.1 Subjectivity
    2. 4.2 Sentence Subjectivity Classification
    3. 4.3 Sentence Sentiment Classification
      1. 4.3.1 Assumption of Sentence Sentiment Classification
      2. 4.3.2 Classification Methods
    4. 4.4 Dealing with Conditional Sentences
    5. 4.5 Dealing with Sarcastic Sentences
    6. 4.6 Cross-Language Subjectivity and Sentiment Classification
    7. 4.7 Using Discourse Information for Sentiment Classification
    8. 4.8 Emotion Classification of Sentences
    9. 4.9 Discussion
  12. 5 Aspect Sentiment Classification
    1. 5.1 Aspect Sentiment Classification
      1. 5.1.1 Supervised Learning
      2. 5.1.2 Lexicon-Based Approach
      3. 5.1.3 Pros and Cons of the Two Approaches
    2. 5.2 Rules of Sentiment Composition
      1. 5.2.1 Sentiment Composition Rules
      2. 5.2.2 DECREASE and INCREASE Expressions
      3. 5.2.3 SMALL_OR_LESS and LARGE_OR_MORE Expressions
      4. 5.2.4 Emotion and Sentiment Intensity
      5. 5.2.5 Senses of Sentiment Words
      6. 5.2.6 Survey of Other Approaches
    3. 5.3 Negation and Sentiment
      1. 5.3.1 Negation Words
      2. 5.3.2 Never
      3. 5.3.3 Some Other Common Sentiment Shifters
      4. 5.3.4 Shifted or Transferred Negations
      5. 5.3.5 Scope of Negations
    4. 5.4 Modality and Sentiment
    5. 5.5 Coordinating Conjunction But
    6. 5.6 Sentiment Words in Non-opinion Contexts
    7. 5.7 Rule Representation
    8. 5.8 Word Sense Disambiguation and Coreference Resolution
    9. 5.9 Summary
  13. 6 Aspect and Entity Extraction
    1. 6.1 Frequency-Based Aspect Extraction
    2. 6.2 Exploiting Syntactic Relations
      1. 6.2.1 Using Opinion and Target Relations
      2. 6.2.2 Using Part-of and Attribute-of Relations
    3. 6.3 Using Supervised Learning
      1. 6.3.1 Hidden Markov Models
      2. 6.3.2 Conditional Random Fields
    4. 6.4 Mapping Implicit Aspects
      1. 6.4.1 Corpus-Based Approach
      2. 6.4.2 Dictionary-Based Approach
    5. 6.5 Grouping Aspects into Categories
    6. 6.6 Exploiting Topic Models
      1. 6.6.1 Latent Dirichlet Allocation
      2. 6.6.2 Using Unsupervised Topic Models
      3. 6.6.3 Using Prior Domain Knowledge in Modeling
      4. 6.6.4 Lifelong Topic Models: Learn as Humans Do
      5. 6.6.5 Using Phrases as Topical Terms
    7. 6.7 Entity Extraction and Resolution
      1. 6.7.1 Problem of Entity Extraction and Resolution
      2. 6.7.2 Entity Extraction
      3. 6.7.3 Entity Linking
      4. 6.7.4 Entity Search and Linking
    8. 6.8 Opinion Holder and Time Extraction
    9. 6.9 Summary
  14. 7 Sentiment Lexicon Generation
    1. 7.1 Dictionary-Based Approach
    2. 7.2 Corpus-Based Approach
      1. 7.2.1 Identifying Sentiment Words from a Corpus
      2. 7.2.2 Dealing with Context-Dependent Sentiment Words
      3. 7.2.3 Lexicon Adaptation
      4. 7.2.4 Some Other Related Work
    3. 7.3 Desirable and Undesirable Facts
    4. 7.4 Summary
  15. 8 Analysis of Comparative Opinions
    1. 8.1 Problem Definition
    2. 8.2 Identify Comparative Sentences
    3. 8.3 Identifying the Preferred Entity Set
    4. 8.4 Special Types of Comparison
      1. 8.4.1 Nonstandard Comparison
      2. 8.4.2 Cross-Type Comparison
      3. 8.4.3 Single-Entity Comparison
      4. 8.4.4 Sentences Involving Compare or Comparison
    5. 8.5 Entity and Aspect Extraction
    6. 8.6 Summary
  16. 9 Opinion Summarization and Search
    1. 9.1 Aspect-Based Opinion Summarization
    2. 9.2 Enhancements to Aspect-Based Summary
    3. 9.3 Contrastive View Summarization
    4. 9.4 Traditional Summarization
    5. 9.5 Summarization of Comparative Opinions
    6. 9.6 Opinion Search
    7. 9.7 Existing Opinion Retrieval Techniques
    8. 9.8 Summary
  17. 10 Analysis of Debates and Comments
    1. 10.1 Recognizing Stances in Debates
    2. 10.2 Modeling Debates/Discussions
      1. 10.2.1 JTE Model
      2. 10.2.2 JTE-R Model: Encoding Reply Relations
      3. 10.2.3 JTE-P Model: Encoding Pair Structures
      4. 10.2.4 Analysis of Tolerance in Online Discussions
    3. 10.3 Modeling Comments
    4. 10.4 Summary
  18. 11 Mining Intentions
    1. 11.1 Problem of Intention Mining
    2. 11.2 Intention Classification
    3. 11.3 Fine-Grained Mining of Intentions
    4. 11.4 Summary
  19. 12 Detecting Fake or Deceptive Opinions
    1. 12.1 Different Types of Spam
      1. 12.1.1 Harmful Fake Reviews
      2. 12.1.2 Types of Spammers and Spamming
      3. 12.1.3 Types of Data, Features, and Detection
      4. 12.1.4 Fake Reviews versus Conventional Lies
    2. 12.2 Supervised Fake Review Detection
    3. 12.3 Supervised Yelp Data Experiment
      1. 12.3.1 Supervised Learning Using Linguistic Features
      2. 12.3.2 Supervised Learning Using Behavioral Features
    4. 12.4 Automated Discovery of Abnormal Patterns
      1. 12.4.1 Class Association Rules
      2. 12.4.2 Unexpectedness of One-Condition Rules
      3. 12.4.3 Unexpectedness of Two-Condition Rules
    5. 12.5 Model-Based Behavioral Analysis
      1. 12.5.1 Spam Detection Based on Atypical Behaviors
      2. 12.5.2 Spam Detection Using Review Graph
      3. 12.5.3 Spam Detection Using Bayesian Models
    6. 12.6 Group Spam Detection
      1. 12.6.1 Group Behavior Features
      2. 12.6.2 Individual Member Behavior Features
    7. 12.7 Identifying Reviewers with Multiple Userids
      1. 12.7.1 Learning in a Similarity Space
      2. 12.7.2 Training Data Preparation
      3. 12.7.3 d-Features and s-Features
      4. 12.7.4 Identifying Userids of the Same Author
    8. 12.8 Exploiting Burstiness in Reviews
    9. 12.9 Some Future Research Directions
    10. 12.10 Summary
  20. 13 Quality of Reviews
    1. 13.1 Quality Prediction as a Regression Problem
    2. 13.2 Other Methods
    3. 13.3 Some New Frontiers
    4. 13.4 Summary
  21. 14 Conclusions
  22. Appendix
  23. Bibliography
  24. Index