• The Journal of Information Systems
• /
• v.28 no.1
• /
• pp.1-16
• /
• 2019

Purpose The purpose of this study is to analyze causal relationship between quantity and quality for deriving ranking model of Online reviews. Thus, we propose implications for deriving the ranking model for retrieving Online reviews more effectively. Design/methodology/approach We collected Online review from Tripadvisor web sites which might be a kind of world-famous tourism web sites. We transformed the natural text reviews to quantified data which consists of quantified positive opinions, quantified negative opinions, quantified modification opinions, reviews lengths and grade scores by using opinion mining technologies in R package. We executed corelation and regression analysis about the data. Findings According to the empirical analysis result, this study confirmed that the review length influenced positive opinion, negative opinion and modification opinion. We also confirmed that negative opinion and modification opinion influenced the grade score.

• Journal of Multimedia Information System
• /
• v.9 no.1
• /
• pp.61-68
• /
• 2022

The importance of online reviews is prevalent as more people access goods or places online and make decisions to visit or purchase. However, such reviews are generally provided by short sentences or mere star ratings; failing to provide a general overview of customer preferences and decision factors. This study explored and broke down restaurant reviews found on Google Maps. After collecting and analyzing 5,427 reviews, we vectorized the importance of words using the TF-IDF. We used a random forest machine learning algorithm to calculate the coefficient of positivity and negativity of words used in reviews. As the result, we were able to build a dictionary of words for positive and negative sentiment using each word's coefficient. We classified words into four major evaluation categories and derived insights into sentiment in each criterion. We believe the dictionary of review words and analyzing the major evaluation categories can help prospective restaurant visitors to read between the lines on restaurant reviews found on the Web.

• Journal of Smart Tourism
• /
• v.3 no.3
• /
• pp.5-15
• /
• 2023

Given the increasing information overload among users of online review websites, understanding the manner in which cognitive costs are reduced and efficient information is made reliable has become increasingly important. This study targets a unique consumer review design and explores how reviews from an external peer-to-peer site embedded in an online travel agency (OTA) website influence subsequent evaluation behaviors. The empirical results indicate that (1) embedded external reviews with a high average valence tend to strengthen the influence of the positive evaluation ratio while diminishing the effect of the review volume, and (2) embedded external reviews with a large variance strengthen the positive effect of the review volume while weakening the effect of the positive evaluation ratio on subsequent positive evaluations. The findings provide practical insights for consumers and online platforms.

• International journal of advanced smart convergence
• /
• v.13 no.2
• /
• pp.44-47
• /
• 2024

With the advancement of Internet technology, online sales and purchases of products have become active. Along with this, the importance of user reviews is also being highlighted. Although user reviews are actively utilized for product sales and purchases, it is difficult to quickly and easily obtain useful information due to the abundance of user reviews. Therefore, prioritizing user reviews is a necessary service for customers that requires careful consideration. Metadata, which contains important information, can be effectively used to prioritize user reviews. However, it is crucial to select and use metadata appropriately according to the purpose. Lean Startup proposes a strategy of repeatedly correcting the problems of ideas or making early transitions to continue trying different approaches. In this paper, we propose a three-step method applying the Lean Startup process to analyze ways to prioritize user reviews using metadata: Build Priority, Measure Priority, Learn Priority.

• ETRI Journal
• /
• v.46 no.4
• /
• pp.633-647
• /
• 2024

In e-commerce platforms, sentiment analysis on an enormous number of user reviews efficiently enhances user satisfaction. In this article, an automated product recommendation system is developed based on machine and deep-learning models. In the initial step, the text data are acquired from the Amazon Product Reviews dataset, which includes 60 000 customer reviews with 14 806 neutral reviews, 19 567 negative reviews, and 25 627 positive reviews. Further, the text data denoising is carried out using techniques such as stop word removal, stemming, segregation, lemmatization, and tokenization. Removing stop-words (duplicate and inconsistent text) and other denoising techniques improves the classification performance and decreases the training time of the model. Next, vectorization is accomplished utilizing the term frequency-inverse document frequency technique, which converts denoised text to numerical vectors for faster code execution. The obtained feature vectors are given to the modified convolutional neural network model for sentiment analysis on e-commerce platforms. The empirical result shows that the proposed model obtained a mean accuracy of 97.40% on the APR dataset.

• KIISE Transactions on Computing Practices
• /
• v.22 no.7
• /
• pp.338-343
• /
• 2016

Today, opinion reviews on the Web are often used as a means of information exchange. As the importance of opinion reviews continues to grow, the number of issues for opinion spam also increases. Even though many research studies on detecting spam reviews have been conducted, some limitations of gold-standard datasets hinder research. Therefore, we introduce a new dataset called "Paraphrased Opinion Spam (POS)" that contains a new type of review spam that imitates truthful reviews. We have noticed that spammers refer to existing truthful reviews to fabricate spam reviews. To create such a seemingly truthful review spam dataset, we asked task participants to paraphrase truthful reviews to create a new deceptive review. The experiment results show that classifying our POS dataset is more difficult than classifying the existing spam datasets since the reviews in our dataset more linguistically look like truthful reviews. Also, training volume has been found to be an important factor for classification model performance.

• Journal of the Korean Society for Library and Information Science
• /
• v.46 no.2
• /
• pp.115-132
• /
• 2012

The purpose of this study is to examine the current status of user-contributed reviews for the Next Generation Library Catalogs, and the potential impact of user reviews available from the external sources, including Amazon.com and GoodReads.com. During the period of February 16th through April 4th, 2012, the number of holding libraries and user-contributed reviews, tags and reading lists of ten selected books were examined from the WorldCat. Also the user-contributed reviews for the same books available from Amazon.com and GoodReads.com were examined, and a case of reviews for one book was analyzed. The result shows that only a few users participated in the WorldCat, and user-contributed reviews were rarely used, when compared with tags or reading lists. Several hundred to thousand user-contributed reviews for the same books were available from Amazon.com and GoodReads.com directly linked with bibliographic records. A case of one book from Amazon.com reveals the possibility of distorting the function of user-contribution. With the introduction of the function of user-contribution, it is expected that its growth rate should be carefully observed and its potential impact on users should be thoroughly and systematically analyzed in the near future.

• Fashion & Textile Research Journal
• /
• v.16 no.1
• /
• pp.79-90
• /
• 2014

With the development of modern society, not only have the Internet and e-commerce been progressed but they also made 'consumption patten' diverse. Despite the internet clothing market growth, there is critical a disadvantage, which is consumers is not able to wear the products presented via online pictures. Thus, pictures on the internet are the only information customers can get, which has caused consciousness on the importance of dealing with 'customer review'. In spite of the fact that 'customer review' has undeniably evolved to be one of customers' essential requisites, the research on this subject is very limited. Until now, the studies on the internet shopping consumers' behavior mostly has to do with the features of 'customer review' such as 'a sense of exaggeration', 'usability', 'duality', 'purity', 'professionalism', 'reliability', and the 'similarity', etc.) Therefore, this study categorizes the characteristics of online shopping reviews to 'the number of reviews', 'the article-length', 'the existence of photos', 'the rewards for reviews', 'the contents of the reviews' and 'the freshness of the reviews' and reviews the impact of an features of 'customers' reviews' affecting the internet shopping sales promotion. Moreover, it is to contribute to the marketing strategies of a shopping mall by analyzing consumers' 'purchasing satisfaction', 'the intention of repurchasing', and 'the factors of viral marketing'.

• Knowledge Management Research
• /
• v.16 no.2
• /
• pp.213-226
• /
• 2015

This study investigates the effects of the volume and valence of the movie reviews on the weekly box-office revenues. Existing literature shows that only the volume of movie reviews influences the box office results, but not valence. However, it has limitations in that it includes only the positivity or negativity ratio of the reviews, not the strength of the valence. In order to overcome such limitations, this study includes the degree of valence. This study used approximately 1.3 million reviews about 300 movies as the sample which was collected from a movie review site in an online portal, that is, movie.naver.com. SPSS was used to test the proposed model. The results of this study show different findings compared to those of the previous studies. First, the volume of movie reviews has been a consistent predictor of the box office success throughout the study periods. Second, the ratio of positive reviews has no impact on the first week's results, but shows significant effects on the box office results during the second week. Third, regarding the degree of positivity or negativity in reviews, the degree of positivity has a significant impact on the box office results only during the first week, while the degree of negativity does not have any significant effects on the results. However, from the second week, the situation is reversed; that is, only the degree of negativity has a significant impact on the box office results, but not the positivity.

• The Journal of Society for e-Business Studies
• /
• v.14 no.4
• /
• pp.19-33
• /
• 2009

With the continuously increasing volume of e-commerce transactions, it is now popular to buy some products and to evaluate them on the World Wide Web. The product reviews are very useful to customers because they can make better decisions based on the indirect experiences obtainable through the reviews. Product Reviews are results expressing customer's sentiments and thus are divided into positive reviews and negative ones. However, as the number of reviews in on-line shopping increases, it is inefficient or sometimes impossible for users to read all the relevant review documents. In this paper, we present a sentiment analysis algorithm for automatically classifying subjective opinions of customer's reviews using opinion mining technology. The proposed algorithm is to focus on product reviews of on-line shopping, and provides summarized results from large product review data by determining whether they are positive or negative. Additionally, this paper introduces an automatic review analysis system implemented based on the proposed algorithm, and also present the experiment results for verifying the efficiency of the algorithm.