• The Journal of Asian Finance, Economics and Business
• /
• 제9권8호
• /
• pp.157-167
• /
• 2022

The global laundry detergents market in 2021 was valued at nearly $121 billion, with consumers being reported as heightening their search for hygienic products capable of fighting viruses. Therefore, the researchers undertook a study to determine how product innovation (PI), product quality (PQ), and product attitude (PA) effects Thai consumers' satisfaction (CS) with their purchase of laundry detergent and fabric softener. After the questionnaire's validity and reliability confirmation, the authors used multi-stage random sampling by region and province in January and February 2022 to collect 520 questionnaires. LISREL 9.10 was used in the CFA and SEM analysis of the six hypotheses, which were determined to be supported. The results showed that all three causal variables positively influenced CS, with a total effect (TE) R² value = 87%. Also, latent variable total effect (TE) values showed that PI was strongest (0.93), then PQ (0.56), and finally, PA (0.54). Therefore, consumer satisfaction is essential in a firm's ongoing development and sustainability in a highly competitive, globalized world. Organizations must develop competitive strategies that adjust to consumer needs. Management must monitor online and social media sources where product reviews are given and adjust their strategies accordingly.

• 경영정보학연구
• /
• 제23권1호
• /
• pp.45-68
• /
• 2021

온라인 리뷰가 소비자 의사결정에 미치는 영향이 증가함에 따라 리뷰조작에 대한 염려도 증가하고 있다. 리뷰조작은 판매량을 증가시키기 위해, 진실 되지 않은 리뷰를 게시하는 것으로 소비자의 역 선택을 초래하며, 사회 전체에 큰 비용으로 작용한다. 선행연구는 대부분 데이터 마이닝 방법을 통해 리뷰조작을 예측하는 데 초점을 맞추었으며, 소비자 관점의 연구는 상대적으로 제한적이다. 그러나 소비자가 지각한 리뷰의 조작 가능성은 리뷰의 유용성에 영향을 미칠 수 있으므로 허위 여부와 상관없이 온라인 구전 관리에 중요한 시사점을 제공할 수 있다. 따라서 본 연구에는 소비자가 조작되었다고 평가한 리뷰와 일반적인 리뷰 간에 어떠한 차이가 있는지 분석하고, 조작된 것으로 평가된 리뷰와 리뷰 유용성 간의 관계를 분석하였다. 실증분석을 위해 LibraryThing 웹사이트의 온라인 도서 리뷰 34,711개를 다수준 로지스틱 회귀분석과 포아송 회귀분석을 활용하여 분석하였다. 분석결과 소비자가 조작되었다고 지각하는 리뷰와 그렇지 않은 리뷰 간에는 제품 수준, 리뷰어 수준, 리뷰 수준 요인들에 차이가 있는 것으로 나타났다. 또한, 조작된 리뷰는 리뷰 유용성에 부정적인 영향을 미치는 것으로 나타났다.

• 한국IT서비스학회지
• /
• 제15권2호
• /
• pp.169-184
• /
• 2016

Smartphone use is rapidly spreading due to the advantage of being able to connect to the Internet anytime, anywhere--and mobile app development is developing accordingly. The characteristic of the mobile app market is the ability to launch one's app into foreign markets with ease as long as the platform is the same. However, a large amount of prior research asserts that consumers behave differently depending on their culture and, from this perspective, various studies comparing the differences between consumer behaviors in different countries exist. Accordingly, this research, which uses online product reviews (OPRs) in order to analyze the cultural differences in consumer behavior comparatively by nationality, proposes to compare the U.S. and South Korea by selecting ten apps which were released in both countries in order to perform a sentimental analysis on the basis of star ratings and, based on those ratings, to interpret the sentiments in reviews. This research was carried out to determine whether, on the basis of ratings analysis, analysis of review contents for sentiment differences, analysis of LDA topic modeling, and co-occurrence analysis, actual differences in online reviews in South Korea and the U.S. exist due to cultural differences. The results confirm that the sentiments of reviews for both countries appear to be more negative than those of star ratings. Furthermore, while no great differences in high-raking review topics between the U.S. and South Korea were revealed through topic modeling and co-occurrence analyses, numerous differences in sentiment appeared-confirming that Koreans evaluated the mobile apps' specialized functions, while Americans evaluated the mobile apps in their entirety. This research reveals that differences in sentiments regarding mobile app reviews due to cultural differences between Koreans and Americans can be seen through sentiment analysis and topic modeling, and, through co-occurrence analysis, that they were able to examine trends in review-writing for each country.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제18권2호
• /
• pp.263-283
• /
• 2024

Users' comments after online shopping are critical to product reputation and business improvement. These comments, sometimes known as e-commerce reviews, influence other customers' purchasing decisions. To confront large amounts of e-commerce reviews, automatic analysis based on machine learning and deep learning draws more and more attention. A core task therein is sentiment analysis. However, the e-commerce reviews exhibit the following characteristics: (1) inconsistency between comment content and the star rating; (2) a large number of unlabeled data, i.e., comments without a star rating, and (3) the data imbalance caused by the sparse negative comments. This paper employs Bidirectional Encoder Representation from Transformers (BERT), one of the best natural language processing models, as the base model. According to the above data characteristics, we propose the F_MixBERT framework, to more effectively use inconsistently low-quality and unlabeled data and resolve the problem of data imbalance. In the framework, the proposed MixBERT incorporates the MixMatch approach into BERT's high-dimensional vectors to train the unlabeled and low-quality data with generated pseudo labels. Meanwhile, data imbalance is resolved by Focal loss, which penalizes the contribution of large-scale data and easily-identifiable data to total loss. Comparative experiments demonstrate that the proposed framework outperforms BERT and MixBERT for sentiment analysis of e-commerce comments.

• 한국정보통신학회:학술대회논문집
• /
• 한국정보통신학회 2019년도 춘계학술대회
• /
• pp.502-503
• /
• 2019

리뷰 데이터 분석을 통해 추천을 하는 기존 시스템에서 사용자의 특성 혹은 상품 구매 취향와 같은 개인의 선호 세부 정보를 반영하지 않는 점을 보완하여 본 논문에서는 사용자가 상품을 검색하고 그 상품을 구매할 때 가장 중요하게 생각하는 기준을 선택하도록 하고, 이를 반영하여 분석함으로써 다양한 사용자에게 맞춤화된 추천 정보를 제공하는 시스템을 제안한다. 이는 사용자가 상품 구매 시 가장 큰 비중을 차지하는 기준을 토대로 가중치를 부여하여 감성분석을 수행하고 그 결과를 반영하여 상품 목록을 제공한다. 따라서, 상품 추천 정보에 사용자 개인의 선호도를 반영하였기 때문에 기존 추천 시스템을 통해 상품을 추천받는 것보다 효율적인 결과를 얻을 수 있을 것으로 사료된다.

• Asia pacific journal of information systems
• /
• 제23권3호
• /
• pp.131-154
• /
• 2013

New internet technologies have created a revolutionary new platform which allows consumers to make decision about product price and quality quickly and provides information about themselves through the transcript of online reviews. By expressing their feelings toward products or services on virtual opinion platforms, users extend their influence into cyberspace as electronic word-of-mouth (e-WOM). Existing research indicates that an impact of eWOM on the consumer decision process is influential. For both academic researchers and practitioners, investigating this phenomenon of information sharing in online website is essential given the increasing number of consumers using them as sources of purchase decisions. It is worthwhile to examine the extent to which opinion seekers are willing to accept and adopt online reviews and which factors encourage adoption. Discerning the most motivating aspects of information adoption in particular, could help electronic marketers better promote their brand and presence on the internet. The objectives of this study are to investigate how online WOM influences a persons' purchase decision by discovering which factors encourage information adoption. Especially focused on the self-efficacy, this research investigates how self-efficacy affects on information usefulness and adoption of online information. Although people are exposed to same review or comment about product or service, some accept the reviews while others do not. We notice that accepting online reviews mainly depends on the person's preference or personal characteristics. This study empirically examines this issue by using cognitive dissonance theory. Specifically, in the movie industry, we address few questions-is always positive WOM generating positive effect? What if the movie isn't the person's favorite genre? What if the person who is very self-assertive so doesn't take other's opinion easily? In these cases of cognitive dissonance, is always WOM generating same result? While many studies have focused on one direct of WOM which indicates positive (or negative) informative reviews or comments generate positive (or negative) results and more (or less) profits, this study investigates not only directional properties of WOM but also how people change their opinion towards product or service positive to negative, negative to positive through the online WOM. An experiment was conducted quantitatively by using a sample of 168 users who have experience within the online movie review site, 'Naver Movie'. Users were required to complete a survey regarding reviews and comments taken from the real movie page. The data reflected user's perceptions of online WOM information that determined users' adoption level. Analysis results provide empirical support for the proposed theoretical perspective. When user can't agree with the opinion of online WOM information, in other words, when cognitive dissonance between online WOM information and users' preference occurs, perceived self-efficacy significantly decreases customers' perception of usefulness. And this perception of usefulness plays an important role in determining users' intention to adopt online WOM information. Most of researches have been concentrated on characteristics of online WOM itself such as quality or vividness of information, credibility of source and direction of online WOM, etc. for describing effect of online WOM, but our results suggest that users' personal character (e.g., self-efficacy) plays decisive role for acceptance of online WOM information. Higher self-efficacy means lower possibility to accept the information that represents counter opinion because of cognitive dissonance, whereas the people that have lower self-efficacy are willing to accept the online WOM information as true and refer to purchase decision. This study suggests a model for understanding role of direction of online WOM information. Also, our result implicates the importance of online review supervision and personalized information service by confirming switching opinion negative to positive is more difficult than positive to negative through the online WOM information. This implication would help marketers to manage online reviews of their products or services.

• 지능정보연구
• /
• 제27권2호
• /
• pp.1-15
• /
• 2021

추천시스템은 사용자의 기호를 파악하여 물품 구매 결정을 도와주는 역할을 할 뿐만 아니라, 비즈니스 전략의 관점에서도 중요한 역할을 하기에 많은 기업과 기관에서 관심을 갖고 있다. 최근에는 다양한 추천시스템 연구 중에서도 NLP와 딥러닝 등을 결합한 하이브리드 추천시스템 연구가 증가하고 있다. NLP를 이용한 감성분석은 사용자 리뷰 데이터가 증가함에 따라 2000년대 중반부터 활용되기 시작하였지만, 기계학습 기반 텍스트 분류를 통해서는 텍스트의 특성을 완전히 고려하기 어렵기 때문에 리뷰의 정보를 식별하기 어려운 단점을 갖고 있다. 본 연구에서는 기계학습의 단점을 보완하기 위하여 BERT 기반 감성분석을 활용한 추천시스템을 제안하고자 한다. 비교 모형은 Naive-CF(collaborative filtering), SVD(singular value decomposition)-CF, MF(matrix factorization)-CF, BPR-MF(Bayesian personalized ranking matrix factorization)-CF, LSTM, CNN-LSTM, GRU(Gated Recurrent Units)를 기반으로 하는 추천 모형이며, 실제 데이터에 대한 분석 결과, BERT를 기반으로 하는 추천시스템의 성과가 가장 우수한 것으로 나타났다.

• 한국게임학회 논문지
• /
• 제17권3호
• /
• pp.33-44
• /
• 2017

인터넷 환경의 발달로 소비자들 사이에 상품정보에 대한 의견이 교환되기 시작하면서 다양한 형식의 온라인 리뷰들이 급속도로 생성되고 있다. 이러한 추세에 따라, 기업들은 온라인 리뷰들을 분석하여 마케팅, 세일즈, 제품개발 등의 다양한 기업 활동에서 그 결과를 활용하려는 노력을 진행하고 있다. 그러나 대표적인 경험재인 '게임'과 관련된 산업에서의 온라인 리뷰에 대한 연구는 매우 부족한 실정이다. 이에 본 연구는 머신러닝 모델을 활용하여 스팀(STEAM)게임의 커뮤니티 데이터를 분석하였다. 이를 통해 타 사용자의 게임 리뷰를 유용하다고 판단하는데 영향을 미치는 요인을 분석하고, 리뷰의 유용성을 예측하는데 있어 가장 우수한 성능을 보인 모델과 변수들을 도출하여 사용자의 충성도와 사용성을 증대시키기 위한 제안을 하고자 한다.

• Journal of Information Science Theory and Practice
• /
• 제3권1호
• /
• pp.6-23
• /
• 2015

This study develops an effective method for sentiment analysis of user-generated content on drug review websites, which has not been investigated extensively compared to other general domains, such as product reviews. A clause-level sentiment analysis algorithm is developed since each sentence can contain multiple clauses discussing multiple aspects of a drug. The method adopts a pure linguistic approach of computing the sentiment orientation (positive, negative, or neutral) of a clause from the prior sentiment scores assigned to words, taking into consideration the grammatical relations and semantic annotation (such as disorder terms) of words in the clause. Experiment results with 2,700 clauses show the effectiveness of the proposed approach, and it performed significantly better than the baseline approaches using a machine learning approach. Various challenging issues were identified and discussed through error analysis. The application of the proposed sentiment analysis approach will be useful not only for patients, but also for drug makers and clinicians to obtain valuable summaries of public opinion. Since sentiment analysis is domain specific, domain knowledge in drug reviews is incorporated into the sentiment analysis algorithm to provide more accurate analysis. In particular, MetaMap is used to map various health and medical terms (such as disease and drug names) to semantic types in the Unified Medical Language System (UMLS) Semantic Network.

• 한국정보통신학회:학술대회논문집
• /
• 한국정보통신학회 2014년도 춘계학술대회
• /
• pp.89-93
• /
• 2014

제품 리뷰들은 잠재적인 고객의 구매 선택에 매우 중요하다. 제품 리뷰들은 또한 제조사들로 하여금 자신들의 제품의 문제점을 찾고 경쟁자들의 비즈니스 정보를 수집하는 데 사용된다. 그러나 어떤 사람들은 가짜 리뷰를 쓰고, 잠재적인 고객들과 제조사들로 하여금 잘못된 선택을 하게 만든다. 따라서 가짜 리뷰 판별은 전자 상거래 사이트에서 주된 문제들 중 하나이다. 서포트 벡터 머신즈(SVM)는 좋은 성능을 보이는 중요한 텍스트 분류 알고리즘이다. 본 논문에서는 온라인 리뷰 스팸을 판별하기 위해 가중치, Karush-Kuhn-Tucker(KKT) 조건의 확장, 그리고 컨벡스 헐(Convex Hull)에 근거한 점증적 알고리즘을 제시한다. 최종적으로 우리는 제시된 알고리즘의 성능을 이론적으로 분석한다.