Publications

Detailed Information

Cross-cultural difference in product preference in consumer review-based text mining methods : 소비자 리뷰 기반 텍스트 마이닝 기법들로 보는 문화 간 제품 선호도 차이: 스마트 밴드 사례 연구
A case study on smart band

DC Field Value Language
dc.contributor.advisor윤명환-
dc.contributor.authorCai Wang-
dc.date.accessioned2020-10-13T02:43:17Z-
dc.date.available2020-10-13T02:43:17Z-
dc.date.issued2020-
dc.identifier.other000000163107-
dc.identifier.urihttps://hdl.handle.net/10371/169186-
dc.identifier.urihttp://dcollection.snu.ac.kr/common/orgView/000000163107ko_KR
dc.description학위논문 (석사) -- 서울대학교 대학원 : 공과대학 산업공학과, 2020. 8. 윤명환 .-
dc.description.abstractThe aim of this study is to prove that the consumer review-based text mining methods proposed in the paper for cross-cultural design are effective. To prove it, we took Mi band 3 as a case study where we compared the cross-cultural differences in product preference of users from different cultural regions with this method.

With the development of global market, more and more products and services are sold across the globe. Users from different cultures have different behaviors, cognitive styles, and value systems. Therefore, product should be designed to meet the needs and preferences of users from different cultural groups. In the field of cross-cultural design, existing studies are mainly focused on traditional usability and UX research methods. However, these methods expose some disadvantages when applied into cross-cultural design contexts.

E-commerce websites provide a large volume of product reviews and it is easy to collect review data online. There is no need to employ foreign participants or make a survey onsite or remotely, which will save much more cost and time. There is a new trend that customer reviews are examined to know consumer opinions. Neverlessness, there are not many studies by analyzing online reviews in the field of cross-cultural design.

Thus, my research proposed consumer review-based text mining methods for cross-cultural design, which consist of aspect-level opinion mining, sentiment analysis, and semantic network analysis.

We collected review data from the following three websites: Naver of South Korea, Jingdong of China, and Amazon of the United States. Text mining methods including opinion mining, sentiment analysis, and semantic network analysis were performed. Firstly, product aspects were extracted from reviews according to word frequency. This indicates how much users are paying attention to different aspects of the product. Aspect-level sentiment analysis was conducted to find out customer satisfaction with different product aspects. Then, the words most associated with each product aspect were listed. Cluster analysis was conducted and the topic of each cluster was summarized. Data visualization of each dataset was done. Lastly, cross-cultural difference among three countries from the results was observed and discussed.

Though there exist similar issues in product preferences of users from South Korea, China, and the United States, cross-cultural differences about Mi band 3 are shown in many product aspects.

Korean tend to take Mi band as a fashionable, cool, yet not useful wearable device. They often buy it as a nice gift. They are interested in the appearance of the strap and often buy straps of different colors and materials. Korean do not enjoy outdoor activities as much as American. And the function of NFC is not prevalent in Korea. Thus, the smart band is not useful to Korean. These can explain why Korean do not care about quality of the smart band and do not want to buy Mi band at a high price.

Korean think that the language of Korean on the display, application, and manual is the most important feature. The length of Korean texts is longer than Chinese to convey the same information. On the other hand, Korean prefer to check message notification on smart band rather than call notification. Therefore, Korean need a larger size for screen.

Chinese are more concerned about different kinds of functions including fitness tracker (step counting, heart rate monitoring, and sleep monitoring), notification, and NFC. These different functions are all important and practical to Chinese.

American enjoy outdoor activities and tend to use smart band mostly as activity tracker. They care more about activity tracker function including heart rate monitoring and step counting than Korean and Chinese. They have a higher requirement about the accuracy of measured data and have more negative reviews on activity tracker function than Korean and Chinese. Besides, they need the mode for swimming. Because American usually use the smart band for outdoor activities, they complain a lot that the screen is prone to scratches and is invisible under the outdoor sunlight. Also, they pay attention to the quality of screen and strap, expecting the material make the screen and strap durable. Besides, battery is the most significant aspect to American. They always try to test each function to find which function makes battery life short.

The results of the case study prove that the consumer review-based text mining method proposed in the paper can generate cross-cultural difference in product preference effectively, which is helpful to cross-cultural design research. And this method is relatively easy and fast compared to other conventional methods.
-
dc.description.tableofcontentsChapter 1. Introduction 1
1.1 Background and Motivation 1
1.2 Research Objective 3
1.3 Organization of the Thesis 4

Chapter 2. Literature Review 5
2.1 Cross-Cultural Design 5
2.1.1 Definition 5
2.1.2 Necessity 6
2.1.3 Method 7
2.2 Opinion Mining and Sentiment Analysis 10
2.2.1 Aspect Level Opinion Mining 10
2.2.2 Cross-Lingual Opinion Mining 11
2.3 Semantic Network Analysis 13

Chapter 3. Methodology 15
3.1 Data Collection 15
3.2 Data Processing 16
3.2.1 Text Preprocessing 16
3.2.2 Opinion Mining and Sentiment Analysis 16
3.2.3 Semantic Network Analysis 17
3.2.4 Result Sample 18

Chapter 4. Result 20
4.1 Overview 20
4.2 Opinion Mining and Sentiment Analysis 21
4.2.1 Normalized Frequency 21
4.2.2 Sentiment Analysis 23
4.3 Semantic Network Analysis 26
4.3.1 Associated Words 26
4.3.1 Cluster Analysis 31
4.3.1 Data Visualization 34
4.4 Results based on Aspects 37
4.4.1 Battery 37
4.4.2 Price 39
4.4.3 Function 41
4.4.4 Step Counting 43
4.4.5 Korean 45
4.4.6 Heart Rate Monitoring 47
4.4.7 Sleep Monitoring 49
4.4.8 Quality 51
4.4.9 Notification 53
4.4.10 Screen 55
4.4.11 Exercise 57
4.4.12 App 59
4.4.13 Call 61
4.4.14 Connection 63
4.4.15 Waterproof 65
4.4.16 Display 67
4.4.17 Message 69
4.4.18 Alarm 71
4.4.19 Gift 73
4.4.20 Strap 75

Chapter 5. Conclusion 78
5.1 Summary of Findings 78
5.2 Future Research 80


Bibliography 82
-
dc.language.isoeng-
dc.publisher서울대학교 대학원-
dc.subjectText mining-
dc.subjectSentiment analysis-
dc.subjectCross-cultural design-
dc.subjectConsumer reviews-
dc.subjectWearable device-
dc.subjectSmart band-
dc.subject.ddc670.42-
dc.titleCross-cultural difference in product preference in consumer review-based text mining methods-
dc.title.alternative소비자 리뷰 기반 텍스트 마이닝 기법들로 보는 문화 간 제품 선호도 차이: 스마트 밴드 사례 연구-
dc.typeThesis-
dc.typeDissertation-
dc.contributor.AlternativeAuthor왕차이-
dc.contributor.department공과대학 산업공학과-
dc.description.degreeMaster-
dc.date.awarded2020-08-
dc.title.subtitleA case study on smart band-
dc.identifier.uciI804:11032-000000163107-
dc.identifier.holdings000000000043▲000000000048▲000000163107▲-
Appears in Collections:
Files in This Item:

Altmetrics

Item View & Download Count

  • mendeley

Items in S-Space are protected by copyright, with all rights reserved, unless otherwise indicated.

Share