Title Page
ABSTRACT
Contents
CHAPTER 1. INTRODUCTION 18
1.1. Introduction 18
1.2. Background of the study 18
1.3. Problem statement and research gaps 20
1.4. Operational definitions of key terms 23
1.4.1. WOM and eWOM 23
1.4.2. Online reviews 24
1.4.3. Rating 25
1.4.4. Consumer perceptions of helpfulness 26
1.4.5. Text summarization 27
1.5. The rationale for the study 29
1.6. Significance of the study 31
1.7. Research questions 33
1.8. Aim and objectives 34
1.9. Research methods 35
1.10. Research outcomes 35
1.11. Structure of the thesis 36
CHAPTER 2. LITERATURE REVIEW 38
2.1. Introduction 38
2.2. EWOM 38
2.2.1. WOM 38
2.2.2. EWOM 41
2.3. Online consumer review 44
2.4. Components affecting the quality of online reviews 47
2.4.1. Rating 48
2.4.2. Online review features 49
2.5. Online review helpfulness 57
2.6. Theoretical background 60
2.6.1. The Theory of Reasoned Action (TRA) 61
2.6.2. The Theory of Planned Behavior (TPB) 63
2.6.3. Stimulus-Organism-Response Theory (SOR) 65
2.6.4. The Elaboration Likelihood Model (ELM) 66
2.6.5. Aproposed theoretical model 68
2.7. Proposed research framework and hypotheses 70
2.7.1. Proposed research framework 70
2.7.2. Hypotheses of review text and review helpfulness 71
2.7.3. Hypotheses of the moderating effect of rating 74
2.8. Summary 75
2.9. Conclusion 76
CHAPTER 3. TOURISM INDUSTRY IN CHINA 77
3.1. Introduction 77
3.1.1. Tourism in China 77
3.1.2. Tourism in Beijing 80
3.2. The overview of the Great Wall 82
3.2.1. Ba Daling Great Wall 82
3.2.2. Mu Tianyu Great Wall 83
3.3. Mu Tianyu and Ba Daling POI Comparision 84
3.4. Online summaries of Mu Tianyu and Ba Daling Great Walls 85
CHAPTER 4. RESEARCH METHODOLOGY 88
4.1. Introduction 88
4.2. Research philosophy 90
4.3. Research approach 93
4.4. Research methodological choice 94
4.5. Research strategies 96
4.6. Time horizon 98
4.7. Research procedures 98
4.8. Social media analytics 100
4.9. Development of key metrics 101
4.9.1. Exogenous variables 102
4.9.2. Endogenous variables 105
4.9.3. Moderating variables 106
4.9.4. Demographic variables 108
4.9.5. Expert review of the variables 108
4.10. Text summarization 108
4.11. Data collection 109
4.11.1. Population 109
4.11.2. Choice of sampling method 111
4.11.3. Sample size 111
4.11.4. Data analytical methods used in the research 112
4.12. Social media analytics (CUP framework) 113
4.12.1. Data capturing 113
4.12.2. Data understanding 119
4.13. Text summarization techniques 125
4.14. Development of measures 127
4.14.1. Measure of rating 127
4.14.2. Measures of semantic features 127
4.14.3. Measures of linguistic features 129
4.14.4. Measure of pictures 130
4.14.5. Measure of helpfulness 131
4.15. The structure of data process 132
4.16. Summary 135
CHAPTER 5. ANALYSIS AND DISCUSSION 136
5.1. Introduction 136
5.2. Selection of appropriate statistical analysis technique 136
5.3. Multiple linear regression analysis 140
5.4. Review-related features' main effects on helpfulness 140
5.4.1. An overview of the analyses 140
5.4.2. An overview of the results 141
5.4.3. Main effect of rating on review helpfulness 142
5.4.4. Main effects of POI attributes and the moderating effect of review rating 143
5.4.5. Main effects of length and the moderating effect of review rating 144
5.4.6. Main effects of readability and the moderating effect of review rating 144
5.4.7. Main effects of pictures and the moderating effect of review rating 145
5.4.8. The Main effect comparison between Badaling and Mutianyu Great Walls 145
5.4.9. Summary 147
5.5. Discussion 148
5.5.1. Review semantic features and review helpfulness 148
5.5.2. Length and review helpfulness 152
5.5.3. Readability and review helpfulness 153
5.5.4. Pictures and review helpfulness 155
5.5.5. Rating's moderating role between review text and review helpfulness 156
5.6. Text summarization 158
CHAPTER 6. IMPLICATIONS AND CONCLUSIONS 161
6.1. Introductions 161
6.2. Theoretical and practical implications 161
6.2.1. Theoretical implications 161
6.2.2. Practical implications 164
6.3. Limitations 170
6.4. Conclusions 171
REFERENCES 173
APPENDICES 188
Table 2-1. A brief overview of the interrelationships of the constructs 75
Table 2-2. The interrelationships of the hypotheses 76
Table 3-1. POIs basic information 85
Table 3-2. "About" from Lonely Planet.com 86
Table 3-3. "About" from TripAdvisor.com 87
Table 4-1. Key features of positivism, constructivist and critical theory research paradigm 92
Table 4-2. Quantitative or qualitative designs 95
Table 4-3. Research strategies 96
Table 4-4. Descriptive and experimental research 97
Table 4-5. Items used to measure constructs 102
Table 4-6. Review numbers for each heritage site 115
Table 4-7. Preprocessing example 118
Table 4-8. English online review topics 124
Table 4-9. Topic grouping 124
Table 4-10. Review rating of Great Walls 127
Table 4-11. Values of attributes 129
Table 4-12. Flesch Reading Ease Formula 130
Table 4-13. The mean values of length and readability 130
Table 4-14. The mean of pictures 130
Table 4-15. The helpfulness scores 132
Table 5-1. Comparative analysis of Mutianyu and Badaling Great Walls 142
Table 5-2. Rating effect. 143
Table 5-3. Hypotheses 1 and 1a testing. 144
Table 5-4. Hypotheses 2 and 2a testing. 144
Table 5-5. Hypotheses 3 and 3a testing. 145
Table 5-6. Hypotheses 4 and 4a testing. 145
Table 5-7. Comparative analysis of Mutianyu and Badaling Great Walls. 147
Table 5-8. Attributes. 147
Table 5-9. Text summarization of Badaling and Mutianyu 160
Figure 1-1. A webpage of an online review 21
Figure 1-2. Research phases 23
Figure 1-3. Rating on an Apple Pencil 26
Figure 1-4. "Popular mentions in TripAdvisor" 28
Figure 1-5. "About" of Tower of London 28
Figure 1-6. Text corpus-based tourism big data mining 29
Figure 2-1. Reviews from TripAdvisor.com 45
Figure 2-2. A check-in example created via Ctrip 47
Figure 2-3. A comparison between an in-depth review and a poor review 53
Figure 2-4. A well-written review with high readability 55
Figure 2-5. A poor-written review with high readability 55
Figure 2-6. Pictures in a review 57
Figure 2-7. "Thumb ups" 60
Figure 2-8. The Theory of Reasoned Action 62
Figure 2-9. The Theory of Planned Behavior 63
Figure 2-10. SOR model of consumer purchase 66
Figure 2-11. Aproposed theoretical model 69
Figure 2-12. Proposed research model 70
Figure 3-1. Border crossings from Mainland China 2001-2023 78
Figure 3-2. Inbound tourism arrivals and growth rate 80
Figure 3-3. Study setting (Great Walls in Beijing) 82
Figure 3-4. Study setting (Badaling Great Wall tourism map) 83
Figure 3-5. Study setting (Mutianyu Great Wall tourism map) 84
Figure 4-1. The diagram of research methodology 89
Figure 4-2. The research "onion" 90
Figure 4-3. Research process 99
Figure 4-4. Social media analytics process 101
Figure 4-5. The basic process of a web crawler 114
Figure 4-6. A relational database 119
Figure 4-7. Schematic of LDA algorithm 120
Figure 4-8. Topic coherence 123
Figure 4-9. The formula of Tf-idf 126
Figure 4-10. The number of images included in reviews 131
Figure 4-11. The structure of data process 132
Figure 5-1. Topic distribution 152