Muutke küpsiste eelistusi

Sentiment Analysis: Mining Opinions, Sentiments, and Emotions 2nd Revised edition [Kõva köide]

4.00/5 (14 hinnangut Goodreads-ist)
(University of Illinois, Chicago)
  • Formaat: Hardback, 448 pages, kõrgus x laius x paksus: 240x159x27 mm, kaal: 780 g, Worked examples or Exercises
  • Sari: Studies in Natural Language Processing
  • Ilmumisaeg: 15-Oct-2020
  • Kirjastus: Cambridge University Press
  • ISBN-10: 1108486371
  • ISBN-13: 9781108486378
Teised raamatud teemal:
Sentiment Analysis: Mining Opinions, Sentiments, and Emotions 2nd Revised editionSuurem pilt
  • Formaat: Hardback, 448 pages, kõrgus x laius x paksus: 240x159x27 mm, kaal: 780 g, Worked examples or Exercises
  • Sari: Studies in Natural Language Processing
  • Ilmumisaeg: 15-Oct-2020
  • Kirjastus: Cambridge University Press
  • ISBN-10: 1108486371
  • ISBN-13: 9781108486378
Teised raamatud teemal:
Püsilink: https://www.kriso.ee/db/9781108486378.html
Märksõnad:
Sentiment analysis is the computational study of people's opinions, sentiments, emotions, moods, and attitudes. This fascinating problem offers numerous research challenges, but promises insight useful to anyone interested in opinion analysis and social media analysis. This comprehensive introduction to the topic takes a natural-language-processing point of view to help readers understand the underlying structure of the problem and the language constructs commonly used to express opinions, sentiments, and emotions. The book covers core areas of sentiment analysis and also includes related topics such as debate analysis, intention mining, and fake-opinion detection. It will be a valuable resource for researchers and practitioners in natural language processing, computer science, management sciences, and the social sciences. In addition to traditional computational methods, this second edition includes recent deep learning methods to analyze and summarize sentiments and opinions, and also new material on emotion and mood analysis techniques, emotion-enhanced dialogues, and multimodal emotion analysis.

Arvustused

'As a whole, this book serves as a useful introduction to sentiment analysis along with in-depth discussions of linguistic phenomena related to sentiments, opinions, and emotions. Although many sentiment analysis methods are based on machine learning as in other NLP [ Natural Language Processing] tasks, sentiment analysis is much more than just a classification or regression problem, because the natural language constructs used to express opinions, sentiments, and emotions are highly sophisticated, including sentiment shift, implicated expression, sarcasm, and so on. Liu has described these issues and problems very clearly. Readers will find this book to be inspiring and it will arouse their interests in sentiment analysis.' Jun Zhao, Chinese Academy of Sciences

Muu info

A comprehensive introduction to computational analysis of sentiments, opinions, emotions, and moods. Now including deep learning methods.
Preface xi
Acknowledgments xvi
1 Introduction
1(17)
1.1 Sentiment Analysis Applications
5(4)
1.2 Sentiment Analysis Research
9(6)
1.2.1 Different Levels of Analysis
9(2)
1.2.2 Sentiment Lexicon and Its Issues
11(1)
1.2.3 Analyzing Debates and Comments
12(1)
1.2.4 Mining Intent
13(1)
1.2.5 Opinion Spam Detection and Quality of Reviews
14(1)
1.3 Sentiment Analysis As Mini-NLP
15(1)
1.4 My Approach to Writing This Book
16(2)
2 The Problem of Sentiment Analysis
18(37)
2.1 Definition of Opinion
19(13)
2.1.1 Opinion Definition
20(1)
2.1.2 Sentiment Target
21(1)
2.1.3 Sentiment of Opinion
22(2)
2.1.4 Opinion Definition Simplified
24(2)
2.1.5 Reason and Qualifier for Opinion
26(2)
2.1.6 Objective and Tasks of Sentiment Analysis
28(4)
2.2 Definition of Opinion Summary
32(2)
2.3 Affect, Emotion, and Mood
34(12)
2.3.1 Affect, Emotion, and Mood in Psychology
35(2)
2.3.2 Emotion
37(3)
2.3.3 Mood
40(1)
2.3.4 Feeling
41(2)
2.3.5 Affect, Emotion, and Mood in Sentiment Analysis
43(3)
2.4 Different Types of Opinions
46(6)
2.4.1 Regular and Comparative Opinions
46(1)
2.4.2 Subjective and Fact-Implied Opinions
47(4)
2.4.3 First-Person and Non-First-Person Opinions
51(1)
2.4.4 Meta-Opinions
52(1)
2.5 Author and Reader Standpoint
52(1)
2.6 Summary
53(2)
3 Document Sentiment Classification
55(34)
3.1 Supervised Sentiment Classification
57(17)
3.1.1 Classification Using Traditional Machine Learning Algorithms
57(9)
3.1.2 Classification Using a Custom Score Function
66(1)
3.1.3 Classification Using Deep Learning
67(3)
3.1.4 Classification Based on Lifelong Learning
70(4)
3.2 Unsupervised Sentiment Classification
74(5)
3.2.1 Classification Using Syntactic Patterns and Web Search
74(2)
3.2.2 Classification Using Sentiment Lexicons
76(3)
3.3 Sentiment Rating Prediction
79(2)
3.4 Cross-Domain Sentiment Classification
81(3)
3.5 Cross-Language Sentiment Classification
84(2)
3.6 Emotion Classification of Documents
86(2)
3.7 Summary
88(1)
4 Sentence Subjectivity and Sentiment Classification
89(26)
4.1 Subjectivity
91(1)
4.2 Sentence Subjectivity Classification
92(4)
4.3 Sentence Sentiment Classification
96(6)
4.3.1 Assumption of Sentence Sentiment Classification
96(1)
4.3.2 Traditional Classification Methods
97(2)
4.3.3 Deep Learning-Based Methods
99(3)
4.4 Dealing with Conditional Sentences
102(2)
4.5 Dealing with Sarcastic Sentences
104(3)
4.6 Cross-Language Subjectivity and Sentiment Classification
107(2)
4.7 Using Discourse Information for Sentiment Classification
109(1)
4.8 Emotion Classification of Sentences
110(2)
4.9 Multimodal Sentiment and Emotion Classification
112(1)
4.10 Summary
113(2)
5 Aspect Sentiment Classification
115(53)
5.1 Aspect Sentiment Classification
116(10)
5.1.1 Supervised Learning
117(4)
5.1.2 Lexicon-Based Approach
121(4)
5.1.3 Pros and Cons of the Two Approaches
125(1)
5.2 Rules of Sentiment Composition
126(20)
5.2.1 Sentiment Composition Rules
128(7)
5.2.2 DECREASE and INCREASE Expressions
135(3)
5.2.3 SMALL_OR_LESS and LARGE_OR_MORE Expressions
138(3)
5.2.4 Emotion and Sentiment Intensity
141(1)
5.2.5 Senses of Sentiment Words
142(2)
5.2.6 Survey of Other Approaches
144(2)
5.3 Negation and Sentiment
146(7)
5.3.1 Negation Words
146(3)
5.3.2 Never
149(2)
5.3.3 Some Other Common Sentiment Shifters
151(1)
5.3.4 Shifted or Transferred Negations
152(1)
5.3.5 Scope of Negations
152(1)
5.4 Modality and Sentiment
153(5)
5.5 Coordinating Conjunction But
158(2)
5.6 Sentiment Words in Non-Opinion Contexts
160(2)
5.7 Rule Representation
162(2)
5.8 Word Sense Disambiguation and Coreference Resolution
164(2)
5.9 Summary
166(2)
6 Aspect and Entity Extraction
168(59)
6.1 Frequency-Based Aspect Extraction
169(2)
6.2 Exploiting Syntactic Relations
171(11)
6.2.1 Using Opinion and Target Relations
172(7)
6.2.2 Using Part-of and Attribute-of Relations
179(3)
6.3 Using Supervised Learning
182(6)
6.3.1 Hidden Markov Model
182(1)
6.3.2 Conditional Random Fields
183(3)
6.3.3 Deep Learning-Based Methods
186(2)
6.4 Mapping Implicit Aspects
188(4)
6.4.1 Corpus-Based Approach
188(1)
6.4.2 Dictionary-Based Approach
189(3)
6.5 Grouping Aspects into Categories
192(2)
6.6 Exploiting Topic Models
194(22)
6.6.1 Latent Dirichlet Allocation
196(3)
6.6.2 Using Unsupervised Topic Models
199(6)
6.6.3 Using Prior Domain Knowledge in Modeling
205(2)
6.6.4 Lifelong Topic Models: Learn As Humans Do
207(4)
6.6.5 Using Phrases As Topical Terms
211(5)
6.7 Entity Extraction and Resolution
216(8)
6.7.1 The Problem of Entity Extraction and Resolution
217(3)
6.7.2 Entity Extraction
220(2)
6.7.3 Entity Linking
222(2)
6.7.4 Entity Search and Linking
224(1)
6.8 Opinion Holder and Time Extraction
224(1)
6.9 Summary
225(2)
7 Sentiment Lexicon Generation
227(16)
7.1 Dictionary-Based Approach
228(4)
7.2 Corpus-Based Approach
232(6)
7.2.1 Identifying Sentiment Words from a Corpus
232(1)
7.2.2 Dealing with Context-Dependent Sentiment Words
233(3)
7.2.3 Lexicon Adaptation
236(1)
7.2.4 Some Other Related Work
237(1)
7.3 Sentiment Word Embedding
238(1)
7.4 Desirable and Undesirable Facts
239(2)
7.5 Summary
241(2)
8 Analysis of Comparative Opinions
243(16)
8.1 Problem Definition
243(4)
8.2 Identifying Comparative Sentences
247(1)
8.3 Identifying the Preferred Entity Set
248(2)
8.4 Special Types of Comparisons
250(7)
8.4.1 Nonstandard Comparisons
250(3)
8.4.2 Cross-Type Comparison
253(1)
8.4.3 Single-Entity Comparison
254(1)
8.4.4 Sentences Involving Compare or Comparison
255(2)
8.5 Entity and Aspect Extraction
257(1)
8.6 Summary
258(1)
9 Opinion Summarization and Search
259(14)
9.1 Aspect-Based Opinion Summarization
260(2)
9.2 Enhancements to Aspect-Based Summaries
262(4)
9.3 Contrastive View Summarization
266(1)
9.4 Traditional Summarization
266(1)
9.5 Summarization of Comparative Opinions
267(1)
9.6 Opinion Search
267(2)
9.7 Existing Opinion Retrieval Techniques
269(2)
9.8 Summary
271(2)
10 Analysis of Debates and Comments
273(21)
10.1 Recognizing Stances in Debates
274(3)
10.2 Modeling Debates/Discussions
277(13)
10.2.1 JTE Model
279(5)
10.2.2 JTE-R Model: Encoding Reply Relations
284(2)
10.2.3 JTE-P Model: Encoding Pair Structures
286(2)
10.2.4 Analysis of Tolerance in Online Discussions
288(2)
10.3 Modeling Comments
290(2)
10.4 Summary
292(2)
11 Mining Intent
294(10)
11.1 The Problem of Intent Mining
294(4)
11.2 Intent Classification
298(3)
11.3 Fine-Grained Mining of Intent
301(1)
11.4 Summary
302(2)
12 Detecting Fake or Deceptive Opinions
304(50)
12.1 Different Types of Spam
307(8)
12.1.1 Harmful Fake Reviews
308(1)
12.1.2 Types of Spammers and Spamming
309(2)
12.1.3 Types of Data, Features, and Detection
311(2)
12.1.4 Fake Reviews versus Conventional Lies
313(2)
12.2 Supervised Fake Review Detection
315(3)
12.3 Supervised Yelp Data Experiment
318(4)
12.3.1 Supervised Learning Using Linguistic Features
319(2)
12.3.2 Supervised Learning Using Behavioral Features
321(1)
12.4 Automated Discovery of Abnormal Patterns
322(7)
12.4.1 Class Association Rules
322(2)
12.4.2 Unexpectedness of One-Condition Rules
324(3)
12.4.3 Unexpectedness of Two-Condition Rules
327(2)
12.5 Model-Based Behavioral Analysis
329(4)
12.5.1 Spam Detection Based on Atypical Behaviors
330(1)
12.5.2 Spam Detection Using Review Graphs
331(1)
12.5.3 Spam Detection Using Bayesian Models
332(1)
12.6 Group Spam Detection
333(8)
12.6.1 Group Behavior Features
337(3)
12.6.2 Individual Member Behavior Features
340(1)
12.7 Identifying Reviewers with Multiple Userids
341(7)
12.7.1 Learning in a Similarity Space
342(1)
12.7.2 Training Data Preparation
343(1)
12.7.3 d-Features and s-Features
344(1)
12.7.4 Identifying Userids of the Same Author
345(3)
12.8 Exploiting Burstiness in Reviews
348(3)
12.9 Future Research Directions
351(1)
12.10 Summary
352(2)
13 Quality of Reviews
354(6)
13.1 Quality Prediction As a Regression Problem
354(2)
13.2 Other Methods
356(2)
13.3 Some New Frontiers
358(1)
13.4 Summary
359(1)
14 Conclusion
360(5)
Appendix 365(11)
Bibliography 376(51)
Index 427
Bing Liu is a distinguished professor of Computer Science at the University of Illinois at Chicago. His current research interests include sentiment analysis, lifelong machine learning, natural language processing, and data mining. He has published extensively in top conferences and journals, and his research has been cited on the front page of the New York Times. Three of his research papers also received Test-of-Time awards. He is the recipient of ACM SIGKDD Innovation Award in 2018, and is a Fellow of the ACM, AAAI, and IEEE. He served as the Chair of ACM SIGKDD from 2013-2017.