• Journal of Korean Society for Quality Management
• /
• v.48 no.3
• /
• pp.421-431
• /
• 2020

Purpose: The purpose of this study is to compare and analyze the tokenizer in natural language processing for customer satisfaction in sentiment analysis. Methods: In this study, a supervised learning-based tokenizer Mecab-Ko and an unsupervised learning-based tokenizer SentencePiece were used for comparison. Three algorithms: Naïve Bayes, k-Nearest Neighbor, and Decision Tree were selected to compare the performance of each tokenizer. For performance comparison, three metrics: accuracy, precision, and recall were used in the study. Results: The results of this study are as follows; Through performance evaluation and verification, it was confirmed that SentencePiece shows better classification performance than Mecab-Ko. In order to confirm the robustness of the derived results, independent t-tests were conducted on the evaluation results for the two types of the tokenizer. As a result of the study, it was confirmed that the classification performance of the SentencePiece tokenizer was high in the k-Nearest Neighbor and Decision Tree algorithms. In addition, the Decision Tree showed slightly higher accuracy among the three classification algorithms. Conclusion: The SentencePiece tokenizer can be used to classify and interpret customer sentiment based on online reviews in Korean more accurately. In addition, it seems that it is possible to give a specific meaning to a short word or a jargon, which is often used by users when evaluating products but is not defined in advance.

• Journal of the Korean Society of Clothing and Textiles
• /
• v.46 no.4
• /
• pp.638-656
• /
• 2022

This study aimed to identify characteristics of Generation MZ's consumer purchase decision journey to develop the new fashion CDJ model. The initial stage was affected by habit, online community, social media, aesthetics, circumstantial need, and proxy. In the search and consideration stage, mobile channels were used actively. In the active search and evaluation stage, online media, experiential data, and personal information were employed. In the purchase stage, zoomers took plenty of time in search and evaluation before spending, contrary to millennials who made their purchases more quickly. In the post-purchase experience stage, zoomers actively displayed follow-up behaviors depending on their satisfaction, such as retaining or deleting the app. While, millennials did not turn away from the store or brand, but followed up on their purchases even when they had an unsatisfactory experience. Based on the characteristics of CDJ, iterative cycle CDJ models were developed. Zoomers CDJ model was presented as a search loop that consists of the search and evaluation process, in which information accumulates, and a purchase loop in which the actual purchase occurs. The iterative cycle CDJ model was presented connected to the loyalty loop as the main section, which is accelerated in millennials' CDJ model.

• Journal of Distribution Science
• /
• v.22 no.7
• /
• pp.1-11
• /
• 2024

Purpose: Shopee, Indonesia's most frequently visited marketplace in November 2023, had 427.2 million visits. Supported by the live streaming feature, Shopee has become the most widely used online shopping platform, with an 83.4% market share in 2022. Several factors, such as brand trust and promotions, have significantly influenced Shopee's dominance and consumer purchasing decisions. This research aims to investigate the effect of cosmetic product distribution, brand trust, and promotions on purchasing decisions, considering gender and age as control variables. Research design, data and methodology: A quantitative approach using a survey research method was employed with a sample of 150 respondents, who were followers of the Shopee ESQA Cosmetics account, obtained through the Yamane formula. Data was collected via an online questionnaire. The data analysis technique used in this study was PLS-SEM with Smart PLS software. The results of this research indicate a significant effect of the distribution of cosmetics products, brand trust, promotions, gender, and age as control variables on the purchase decision variable. Conclusions: The distribution of cosmetic products, brand trust and promotions have a positive and significant impact on purchase decisions during live streaming on Shopee, and control variables (gender and age 36-45) have a positive and significant influence on purchase decisions.

• Journal of Engineering Education Research
• /
• v.26 no.3
• /
• pp.20-34
• /
• 2023

This research aims to explore the factors affecting engineering students' major persistence and career decision. To achieve this goal, the experiences and perceptions of engineering students and other stakeholders regarding the factors influencing engineering students' major persistence and career decision were analyzed through online focus group interviews (FGI). The analysis of the interviews resulted in the identification of seven categories, which include meeting with older alumni, team project experiences, and internship experiences as factors influencing major persistence. Additionally, engineering interest, engineering self-efficacy, employment rates, economic factors, and limitations in women engineers' career development were identified as factors affecting career decision. Based on these findings, it is expected that engineering education will develop plans that focus on supporting the major persistence and career development of engineering students, thereby contributing to the future direction of engineering education.

• KDI Journal of Economic Policy
• /
• v.39 no.2
• /
• pp.53-74
• /
• 2017

This paper focuses on the recent extraordinary growth of Chinese cross-border online shopping and draws implications for firm strategies and government policies in Korea to utilize the phenomenon as an opportunity to expand into a broader market via e-commerce exports. I conduct a survey of Chinese cross-border online consumers to identify constraining and determining factors during the stages of their purchase decisions of Korean products. Given the fact that Chinese cross-border online shopping is at the incipient stage and consumers have expressed a strong intent to repurchase, future strategies should focus on attracting new consumers. Accordingly, Korean firms should build a powerful brand image, improve product quality and post-purchase services, and take full advantage of the popularity of the Korean Wave. Meanwhile, the government must step up policy efforts by, for instance, improving e-commerce export statistics, simplifying logistics and clearance procedures, and building trust in Chinese consumers.

• Journal of Korean Society for Quality Management
• /
• v.37 no.1
• /
• pp.40-55
• /
• 2009

This research is an empirical study to analyze the factors that affect users' acceptance & diffusion of online product message. Today many people use the product message/information on the internet, before they make decision to buy something in online/offline environment. This paper proposes the site design look, enjoyment, and usefulness as message perception factors. And we examine the relationship among the perception factors, message credibility, acceptance & diffusion. The purpose of this paper is threefold. First, we examine the relationship between as perception factors and the credibility assessment of the message by customers. Second, we will show the effect of message credibility on the message acceptance & diffusion. Third, we try to present some theoretical and empirical implication through the result of our studies.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.4 no.4
• /
• pp.547-559
• /
• 2010

As web-based online communities are rapidly growing, the agents in the communities need to know their measurable belief of trust for safe and successful interactions. In this paper, we propose a computational model of trust resulting from available feedbacks in online communities. The notion of trust can be defined as an aggregation of consensus given a set of past interactions. The average trust of an agent further represents the center of gravity of the distribution of its trustworthiness and untrustworthiness. Furthermore, we precisely describe the relationships among reputation, trust and average trust through concrete examples showing their computations. We apply our trust model to online social networks in order to show how trust mechanisms are involved in the rational purchasing decision-making of buyers and sellers, and we summarize our simulation results.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• v.15 no.3
• /
• pp.145-152
• /
• 2015

We consider an online selective-sample learning problem for sequence classification, where the goal is to learn a predictive model using a stream of data samples whose class labels can be selectively queried by the algorithm. Given that there is a limit to the total number of queries permitted, the key issue is choosing the most informative and salient samples for their class labels to be queried. Recently, several aggressive selective-sample algorithms have been proposed under a linear model for static (non-sequential) binary classification. We extend the idea to hidden Markov models for multi-class sequence classification by introducing reasonable measures for the novelty and prediction confidence of the incoming sample with respect to the current model, on which the query decision is based. For several sequence classification datasets/tasks in online learning setups, we demonstrate the effectiveness of the proposed approach.

• The Research Journal of the Costume Culture
• /
• v.24 no.6
• /
• pp.803-816
• /
• 2016

The purposes of this study were to investigate the influences of channel assessments on the usage of multi-channels by product types, and the differences in the usage of multi-channels among product types in buying decision making process for fashion products. Data were collected from 510 consumers in their 20s to 50s with purchasing experiences through multi-channel distribution system and living in Seoul and Kyunggi province; 491 were analyzed after deleting incomplete questionnaires. Factor analysis, multiple regression analysis and one-way ANOVA were used for statistical analysis by using SPSS 18.0. The results were as follows: 5 factors were extracted for channel assessment: utility, accuracy, risk, price benefit and sharing information. Price benefits, utility and sharing information for online channel tended to influence positively on the usage of online channel and online+offline channels. Accuracy and low perceived risk of offline influenced positively on offline and on+offline channel usages. The usage levels of on-line and off-line channels for cosmetics were significantly lower than the usage levels for clothes and accessories on information search, evaluation of alternatives, and purchase stages. Significant differences were also found in the usage levels of multi-channels (on+off-line) on information search and evaluation of alternatives stages. The usage levels of the multi-channels for clothes were the highest followed by those of accessories and cosmetics in order.

• Asia pacific journal of information systems
• /
• v.16 no.3
• /
• pp.67-93
• /
• 2006

Some consumers prefer online and others prefer offline. What makes them prefer online or offline? There has been a lack of theoretical development to adequately explain consumers' channel switching behavior between traditional physical stores and new virtual stores. Through consumers' purchase decision processes, this study examined the reasons why consumers changed channels depending on purchase process stages. Consumer's purchase decision process could be divided into three stages: pre-purchase stage, purchase stage, and post-purchase stage. We used the intention of channel selection as a surrogate dependent variable of channel selection. And some constructs, that is, channel function, channel benefits, customer relationship benefits, and perceived behavioral control, were selected as independent variables. In buying look-and-feel products, it was identified that consumers preferred virtual stores to physical stores at pre-purchase stage. To put it concretely, all constructs except channel benefits were more influenced to consumers at virtual stores. This result implied that information searching function, which is a main function at pre-purchase stage, was better supported by virtual stores than physical stores. In purchase stage, consumers preferred physical stores to virtual stores. Specially, all constructs influenced much more to consumers at physical stores. This result implied that although escrow service and trusted third parties were introduced, consumers felt that financial risk, performance risk, social risk, etc. still remained highly online. Finally, consumers did not prefer any channel at post-purchase stage. But three independent variables, i.e. channel function, channel benefits, and customer relationship benefits, were significantly preferred at physical stores rather than virtual stores at post-purchase stage. So we concluded that physical stores were a little more preferred to virtual stores at post-purchase stage. Through this study, it was identified that most consumers might switch channels according to purchase process stages. So, first of all, sales representatives should decide that what benefits should be given them through virtual stores at the pre-purchase stage and through physical stores at the purchase and post-purchase stages, and then devise collaborative channel strategies.