Sentiment In Japanese: A Corpus-based Approach with Sociolinguistic and Cross-lingual Implications Open Access

Abstract

Great progress has been made on sentiment analysis techniques for
in the English language; however, for other languages, sentiment
analysis is less well understood. This thesis reports on statistical
analysis of sentiment in Japanese and English text. Salient features
for each are analyzed to better understand how authors convey
sentiment. In particular, socio-psychological and linguistic
explanations are given for their usage. In addition to proposing
linguistic insights and hypothesis, the foundation is laid for more
effective automatic sentiment classification for non-English
languages.

Table of Contents

1 Introduction 1

Approach 4

Methodology 4

Contribution 8

2 Related Work 9

3 POLAR TOKENS IN JAPANESE REVIEWS 14

3.1 Explicit Modifiers 20

3.2 Scope Extension vs. Scope Restriction in Polar Cases 23

3.3 Sentiment Polarity and Distinctions in Restriction 26

3.4 Statements of Fact and "Explanatory" Statements 29

3.5 The Sentence-final Particle Yo 46

3.6 Opinion Transference with Personal Pronouns 50

3.7 Interrogatives 54

3.8 A Few Words Regarding Politeness 58

3.9 Section 3 Conclusion 60

4 SELECTED TOKENS IN NEUTRAL REVIEWS 61

4.1 Two Sides of a Coin: Contrastive Conjunctions 64

4.2 Explicit Subjectivity and Uncertainty 71

4.3 Properties and Explicit Explanation 73

4.4 Section 4 Conclusion 78

5 COMPARISON TO TRENDS IN ENGLISH 79

5.1 English and Japanese Polar Clauses 80

5.2 Neutral English and Japanese Characteristics 89

5.3 English-Japanese Comparison Summary 97

6 A Classification Experiment 98

7 Conclusion 101

Appendix A: Unedited Japanese Tables 103

Appendix B: A Brief Overview of Basic Japanese 109

Transliteration 111

Pronunciation 111

Brief Grammar Overview 114

Grammar : Equivalence 114

Politeness Levels 115

Adjectives 116

Tenses 116

Negation 117

Questions 117

Bibliography 118

About this Master's Thesis

Rights statement

• Permission granted by the author to include this thesis or dissertation in this repository. All rights reserved by the author. Please contact the author for information regarding the reproduction and use of this thesis or dissertation.

School	Laney Graduate School
Department	Math and Computer Science
Degree	MS
Submission	Master's Thesis
Language	English
Research Field	Computer Science Language, Linguistics
Keyword	sociolinguistics computational linguistics linguistics sentiment analysis cognitive science japanese multilingual sentiment analysis cross-lingual sentiment analysis japanese sentiment analysis subjectivity analysis japanese particles
Committee Chair / Thesis Advisor	Agichtein, Eugene, Emory University
Committee Members	Lu, James, Emory University Wolff, Phillip, Emory University

Last modified

Primary PDF

Thumbnail	Title	Date Uploaded	Actions
	Sentiment In Japanese: A Corpus-based Approach with Sociolinguistic and Cross-lingual Implications ()	2018-08-28 11:39:42 -0400	Press to Select an action Download

Supplemental Files

Thumbnail	Title	Date Uploaded	Actions