User-generated Reviews Research Articles

The eWOM adoption model in the hospitality industry: the moderating effect of the vlogger’s review

酒店业的网络口碑采用意向模型：视频博主评论的调节作用研究目的Vlogger（视频博主）已成为社交媒体营销中的一种新现象, 尤其是在酒店行业, 用户生成的评论可以强烈影响购买决策。 因此, 本文旨在建立一个电子口碑 (eWOM) 采用意向模型（EAM）, 该模型包含来源评估属性（信息质量、主要影响力和来源可信度）、对电子口碑的信任、eWOM intention (EWOMI) 和预订意愿, 以及 (2) 调查视频博主对 EAM 中关系的负面评论的调节。研究设计/方法/途径当前的研究进行了两项研究（ns1 = 446, ns2 = 374）。 数据是通过带有实验设计的在线和面对面调查收集的。 本文使用 CB-SEM 来检查主要效果, 并应用多组分析来测试 vlogger 评论的调节效果。研究发现来源评估的三个属性是网络口碑信任度的重要预测指标, 对网络口碑和预订意愿产生积极影响。 EWOMI 也对预订意愿产生积极影响。 此外, vlogger 的负面评论可以降低信息质量对信任的影响, 以及信任对研究 1 中的 EWOMI 和研究 2 中的酒店预订意愿的影响。研究原创性/价值这项研究为酒店环境建立了 EAM 模型, 并提供了关于视频博主负面评论对 EAM 中关系的调节作用的新颖见解。

Mining the Influencing Factors and Their Asymmetrical Effects of mHealth Sleep App User Satisfaction From Real-world User-Generated Reviews: Content Analysis and Topic Modeling.

Sleep disorders are a global challenge, affecting a quarter of the global population. Mobile health (mHealth) sleep apps are a potential solution, but 25% of users stop using them after a single use. User satisfaction had a significant impact on continued use intention. This China-US comparison study aimed to mine the topics discussed in user-generated reviews of mHealth sleep apps, assess the effects of the topics on user satisfaction and dissatisfaction with these apps, and provide suggestions for improving users' intentions to continue using mHealth sleep apps. An unsupervised clustering technique was used to identify the topics discussed in user reviews of mHealth sleep apps. On the basis of the two-factor theory, the Tobit model was used to explore the effect of each topic on user satisfaction and dissatisfaction, and differences in the effects were analyzed using the Wald test. A total of 488,071 user reviews of 10 mainstream sleep apps were collected, including 267,589 (54.8%) American user reviews and 220,482 (45.2%) Chinese user reviews. The user satisfaction rates of sleep apps were poor (China: 56.58% vs the United States: 45.87%). We identified 14 topics in the user-generated reviews for each country. In the Chinese data, 13 topics had a significant effect on the positive deviation (PD) and negative deviation (ND) of user satisfaction. The 2 variables (PD and ND) were defined by the difference between the user rating and the overall rating of the app in the app store. Among these topics, the app's sound recording function (β=1.026; P=.004) had the largest positive effect on the PD of user satisfaction, and the topic with the largest positive effect on the ND of user satisfaction was the sleep improvement effect of the app (β=1.185; P<.001). In the American data, all 14 topics had a significant effect on the PD and ND of user satisfaction. Among these, the topic with the largest positive effect on the ND of user satisfaction was the app's sleep promotion effect (β=1.389; P<.001), whereas the app's sleep improvement effect (β=1.168; P<.001) had the largest positive effect on the PD of user satisfaction. The Wald test showed that there were significant differences in the PD and ND models of user satisfaction in both countries (all P<.05), indicating that the influencing factors of user satisfaction with mHealth sleep apps were asymmetrical. Using the China-US comparison, hygiene factors (ie, stability, compatibility, cost, and sleep monitoring function) and 2 motivation factors (ie, sleep suggestion function and sleep promotion effects) of sleep apps were identified. By distinguishing between the hygiene and motivation factors, the use of sleep apps in the real world can be effectively promoted.

A Model for Brand Personality Word Embedding: Identifying UNESCO World Heritage Personality Categories

Despite having received considerable academic attention, existing brand personality (BP) scales are difficult to generalize and likely need further development. The aim of this study was to design a novel method for constructing context driven-BP categories through the use of a multi-disciplinary approach and advances in personality studies in psychology and natural language processing. Based on a textual analysis that relies on devising hypotheses of psycholexical representation and the distributed representation of words, the aforementioned method was employed to explore World Heritage Site (WHS) personalities using information from all 1,121 UNESCO World Heritage Sites (WHSs) and 9,920 user-generated reviews on TripAdvisor. The study identified a 192-item WHS personality dictionary organized into five clusters: Exceptionality; Attractiveness; Identification; Responsibility; and Prominence. These categories could be employed to measure other tourism attractions. The results show that UNESCO strongly associates WHSs with the attributes of Attractiveness and Identification.

Improving Recommendation for Effective Personalization in Context-Aware Data Using Novel Neural Network

The digital technologies that run based on users’ content provide a platform for users to help air their opinions on various aspects of a particular subject or product. The recommendation agents play a crucial role in personalizing the needs of individual users. Therefore, it is essential to improve the user experience. The recommender system focuses on recommending a set of items to a user to help the decision-making process and is prevalent across e-commerce and media websites. In Context-Aware Recommender Systems (CARS), several influential and contextual variables are identified to provide an effective recommendation. A substantial trade-off is applied in context to achieve the proper accuracy and coverage required for a collaborative recommendation. The CARS will generate more recommendations utilizing adapting them to a certain contextual situation of users. However, the key issue is how contextual information is used to create good and intelligent recommender systems. This paper proposes an Artificial Neural Network (ANN) to achieve contextual recommendations based on user-generated reviews. The ability of ANNs to learn events and make decisions based on similar events makes it effective for personalized recommendations in CARS. Thus, the most appropriate contexts in which a user should choose an item or service are achieved. This work converts every label set into a Multi-Label Classification (MLC) problem to enhance recommendations. Experimental results show that the proposed ANN performs better in the Binary Relevance (BR) Instance-Based Classifier, the BR Decision Tree, and the Multi-label SVM for Trip Advisor and LDOS-CoMoDa Dataset. Furthermore, the accuracy of the proposed ANN achieves better results by 1.1% to 6.1% compared to other existing methods.

Classifying Scuba Diving Sites through Diver Reviews with a Web Scraping Based UTADIS Application

This research addresses the utilization of user-defined and web-based data for decision-making in recreational activities, focusing on the case of scuba diving, a globally significant recreational pursuit. While a wealth of user-generated information is available on the internet for various leisure activities, harnessing this data involves extensive data collection and organization efforts. Our proposed methodology involves the collection and organization of user reviews, both in verbal and quantitative forms, to create a pertinent dataset for applying multiple criteria decision methodologies to classify diving sites worldwide. An initial dataset containing over 14,000 diving sites worldwide is aggregated into 721 regions and these regions are classified using UTADIS methodology. The research showcases how user-generated review data can be transformed into valuable information, applying classification algorithms of multiple criteria decision analysis within the context of scuba diving. Furthermore, the proposed approach in this research holds the potential to serve as a model for leveraging user-generated data in decision-making processes across various service sectors such as hospitality and leisure that highly rely on customer experience by providing new insights on how more data-driven approaches can be utilized.

Visual Enhancement Capsule Network for Aspect-based Multimodal Sentiment Analysis

Multimodal sentiment analysis, which aims to recognize the emotions expressed in multimodal data, has attracted extensive attention in both academia and industry. However, most of the current studies on user-generated reviews classify the overall sentiments of reviews and hardly consider the aspects of user expression. In addition, user-generated reviews on social media are usually dominated by short texts expressing opinions, sometimes attached with images to complement or enhance the emotion. Based on this observation, we propose a visual enhancement capsule network (VECapsNet) based on multimodal fusion for the task of aspect-based sentiment analysis. Firstly, an adaptive mask memory capsule network is designed to extract the local clustering information from opinion text. Then, an aspect-guided visual attention mechanism is constructed to obtain the image information related to the aspect phrases. Finally, a multimodal fusion module based on interactive learning is presented for multimodal sentiment classification, which takes the aspect phrases as the query vectors to continuously capture the multimodal features correlated to the affective entities in multi-round iterative learning. Otherwise, due to the limited number of multimodal aspect-based sentiment review datasets at present, we build a large-scale multimodal aspect-based sentiment dataset of Chinese restaurant reviews, called MTCom. The extensive experiments both on the single-modal and multimodal datasets demonstrate that our model can better capture the local aspect-based sentiment features and is more applicable for general multimodal user reviews than existing methods. The experimental results verify the effectiveness of our proposed VECapsNet.

Factors affecting user experience of contact tracing app during COVID-19: an aspect-based sentiment analysis of user-generated review

ABSTRACT This study aims to identify the critical factors influencing the user experience of contact tracing apps and the sentiments around them. For this purpose, we used Google play reviews of Aarogya Setu, a contact tracing app developed in India. First, we establish the relationship between review sentiment and review rating using regression between sentiment polarity and review rating. Then, we used a hybrid aspect-based sentiment analysis approach that uses unsupervised linguistic techniques to determine statistically significant concepts present in the review texts and cluster them into representative aspects that were then tagged under human supervision. Finally, supervised deep learning methods were applied for exhaustive extraction of the aspects and associated sentiments from the reviews. The final exercise of determining the key influencing factors was done by grouping these aspects under factors identified by marketing experts. A total of nine factors were identified, with the usefulness of the app being the most important factor. The findings of this study are essential for the development team and government to improve the application and increase adoption.

The Machine Learning Approach for Recommendation System Based on Online Review

For the past few years, various recommendation systems have been introduced that give recommendations and suggestions based on online user reviews. These recommenders provide valuable information from user-generated reviews through various processes. These systems use different techniques to extract the important information from the reviews that are valuable enough to give as suggestions to the users based on their interests. In this research, a portable recommender system has been developed based on the three Machine learning that recommend its user about different mobile phones available in the market and helps them select the best one based on the online user reviews.

Deep Recommendation With Adversarial Training

Recommender systems provide an effective solution to the information overload and have become a research hotspot in both industry and academia. Some existing work on implicit feedback has made great progress. However, there are still the following shortcomings: 1) existing pairwise ranking methods mostly sample unobserved items uniformly to obtain negative samples, which brings a biased solution and insufficient convergence; 2) the recommendation result comes from a complicated process, which makes it less interpretable. Therefore, we put forward a Deep Recommendation with Adversarial Training (DRAT) model, which provides users with personalized recommendations by utilizing an encoder-decoder structure and adversarial training. It consists of two components: one is feature learning module in which the encoder captures the text features of items from user-generated reviews, and the decoder reconstructs the text with the captured features; the other is rating prediction module which utilizes adversarial training to generate personalized negative samples tackling the drawbacks of uniform negative sampling. Extensive experiments on five real-world datasets show that our proposed model significantly outperforms the baseline methods.

Fine-tuning machine translation quality-rating scales for new digital genres: The case of user-generated content

With the active participation of users in product review platforms, online consumer-generated content, and, more specifically, user-generated reviews, have become a clear reference in purchasing decision-making processes, which sometimes exceed the impact of advertising campaigns. A common feature of most tourism review platforms is the use of machine translation (MT) systems to immediately make reviews available to users in various languages. However, the quality of the MT output of these reviews varies greatly, primarily due to the subjective and unstructured nature of this digital genre. Different studies confirm that there are no universal quality rating scales. The assessment of MT output quality usually depends on factors such as the purpose of the text or the value given to the immediacy of the translation. New neural MT systems have been a revolution in the quality increase of the translated output; however, new lines of research are opening up to verify whether the quality of this new paradigm of MT can be assessed with the existing scales, mainly from previous rule-based systems and statistical translation, or whether it is necessary to develop new quality metrics specifically for these new intelligent systems. On the other hand, one of the questions that remain to be resolved in this new context of neural MT is whether the use of large amounts of textual data in the training of these systems is as effective as the use of less data but of higher quality and better-adjusted to the specialty and type of text for which it is used. Based on the hypothesis that each genre requires specific quality rating scales, this work identifies the error patterns and textual characteristics of online user reviews from a corpus-based approach analysis that will contribute to adapting quality rating scales to this specific digital genre.

User-generated reviews and the financial performance of restaurants

PurposeThis paper aims to examine the impact of selected characteristics (rating, volume and variability) of online user-generated reviews on the bottom-line profitability of restaurants.Design/methodology/approachRestaurant-level review data are extracted from TripAdvisor and matched with firm-level data from the financial reports gathered from the Belfirst database of Bureau van Dijk. The resulting sample contains data on 2,297 Belgian firms over the period 2007–2018, for which 134,831 reviews are investigated. The author’s regression model of firm-level profitability is estimated against online review characteristics and various financial control variables, including past profitability. This research model and estimation technique address the endogeneity concerns that typically weaken this kind of study.FindingsWhile comparable studies on hotels document a positive association between review characteristics and profitability, the authors find no relationship between review rating, volume and variability in the profitability of restaurants.Research limitations/implicationsDue to the format of the financial reports of small and medium-sized enterprises (SMEs), data on turnover and cost of materials/services was not available for most restaurants in the sample, limiting our potential for analysis. In addition, our assessment of electronic word of mouth (eWOM) was limited to measures derived from user-generated reviews on TripAdvisor.Practical implicationsIn the literature on eWOM, the importance of online reputation is hardly disputed, especially in the context of the hospitality sector. However, most research to date has focused on the hotel sector and top-line measures of success. This study uses restaurant-level financial data, focuses on bottom-line profitability, considers potential endogeneity issues and pays careful attention to the estimation technique. The results fail to establish a direct relationship between eWOM metrics and financial performance and are surprising, meriting further investigation to establish the underlying causes.Originality/valueIn contrast to prior studies on the impact of eWOM on restaurant performance at a group level, this study examines the impact on unit-level profitability, taking into account several potential sources of estimation bias. In addition, the authors challenge this finding with a battery of sensitivity tests, revalidating the absence of a relationship in each case.

The dissemination and impacts of deceptive eWOM: a dynamic process perspective

ABSTRACT To minimise risk, consumers seek out eWOM, recommendations, and reviews from others before making purchasing decisions. However, it is not uncommon to see companies masterminding deceptive eWOM, disguising promotional messages as user-generated product reviews to attract and retain consumers. The extant literature is extremely limited regarding deceptive eWOM and its long-term effects. Ours is the first longitudinal study to examine consumers’ belief and attitude changes in the context of deceptive eWOM. This study found that, in a herding situation, deceptive eWOM can strongly influence changes in consumers’ beliefs and attitudes, whether in the early phase or mainstream phase, at the pre-adoption stage. If deceptive eWOM emerges earlier, it has a more powerful influence later. At the post-adoption stage, once they have used it, consumers may change their evaluations of a new product. However, this new evaluation is still subject to the initial anchoring effect brought by deceptive eWOM. If consumers sense deceptive eWOM and attribute it to a certain responsible party, the anchoring effect will be weakened. These findings provide further knowledge and understanding of the effects of deceptive eWOM, with implications for firms seeking to develop more effective eWOM marketing strategies.

Does Michelin effect exist? An empirical study on the effects of Michelin stars

PurposeReceiving Michelin stars is considered an effective marketing tool and a prestigious certification in the restaurant industry. However, the direct effects of Michelin stars on restaurant performance remain unclear. To bridge this gap, this study aims to empirically validate the “Michelin effect” on the consumption values of restaurant patrons before and after receiving Michelin stars.Design/methodology/approachThe data, collected from OpenTable, consist of over 160,000 reviews written for 218 restaurants, including 109 Michelin-starred (treatment group) and 109 nonstarred restaurants (control group). The authors measure perceived consumption value using the collected user-generated review data. The authors estimate fixed-effect difference-in-differences regressions to validate the Michelin effect.FindingsMichelin stars enhance social, hedonic and service quality values, which are nonfunctional values. However, no significant effects on functional consumption values, such as economic, food quality and ambience quality values, are observed, even though Michelin stars are pure awards for kitchen performance.Practical implicationsMichelin stars can be an effective marketing tool for fine-dining restaurants because customers consider emotional and nonfunctional benefits such as hedonic and social values, which are more important than functional benefits such as food value. Furthermore, Michelin effects are heterogeneous depending on the number of stars awarded, price range and customers’ gastronomic involvement. These offer a strong rationale for monitoring social media, which may help managers better understand their customers and improve their performance.Originality/valueThis study extends the current literature on the Michelin effect by quantifying consumption values using user-generated review data. To the best of the authors’ knowledge, this study offers the first empirical evidence that directly validates the Michelin effect.

Investigating Consumers’ Online Restaurant Selection Behaviors Using Eye-tracking Technology and Retrospective Think-aloud Interviews

ABSTRACT The purpose of this study was to explore consumers’ attention patterns and information search behaviors when they make online restaurant decisions on consumer review websites. Eye-tracking experiments and retrospective think-aloud interviews were conducted with 30 participants. In the browsing stage, the company-generated advertisements attracted most attention from consumers on one website; however, the attracted attention was not impactful on their dining decisions. In the deliberation stage, instead of reading the user-generated reviews in details, consumers skimmed through reviews; and their attention was attracted more to images embedded within the reviews. In addition, participants verbalized their overall preferences of information that helped them make quick decisions such as images and filter. This study provides valuable insights for hospitality researchers and practitioners.

Unsupervised and Supervised Methods to Estimate Temporal-Aware Contradictions in Online Course Reviews

The analysis of user-generated content on the Internet has become increasingly popular for a wide variety of applications. One particular type of content is represented by the user reviews for programs, multimedia, products, and so on. Investigating the opinion contained by reviews may help in following the evolution of the reviewed items and thus in improving their quality. Detecting contradictory opinions in reviews is crucial when evaluating the quality of the respective resource. This article aims to estimate the contradiction intensity (strength) in the context of online courses (MOOC). This estimation was based on review ratings and on sentiment polarity in the comments, with respect to specific aspects, such as “lecturer”, “presentation”, etc. Between course sessions, users stop reviewing, and also, the course contents may evolve. Thus, the reviews are time dependent, and this is why they should be considered grouped by the course sessions. Having this in mind, the contribution of this paper is threefold: (a) defining the notion of subjective contradiction around specific aspects and then estimating its intensity based on sentiment polarity, review ratings, and temporality; (b) developing a dataset to evaluate the contradiction intensity measure, which was annotated based on a user study; (c) comparing our unsupervised method with supervised methods with automatic feature selection, over the dataset. The dataset collected from coursera.org is in English. It includes 2244 courses and 73,873 user-generated reviews of those courses.The results proved that the standard deviation of the ratings, the standard deviation of the polarities, and the number of reviews are suitable features for predicting the contradiction intensity classes. Among the supervised methods, the J48 decision trees algorithm yielded the best performance, compared to the naive Bayes model and the SVM model.

Heterogeneous Demand Effects of Recommendation Strategies in a Mobile Application: Evidence from Econometric Models and Machine-Learning Instruments

In this paper, we examine the effectiveness of various recommendation strategies in the mobile channel and their impact on consumers’ utility and demand levels for individual products. We find significant differences in effectiveness among various recommendation strategies. Interestingly, recommendation strategies that directly embed social proofs for the recommended alternatives outperform other recommendations. In addition, recommendation strategies combining social proofs with higher levels of induced awareness due to the prescribed temporal diversity have an even stronger effect on the mobile channel. We also examine the heterogeneity of the demand effect across items, users, and contextual settings, further verifying empirically the aforementioned information and persuasion mechanisms and generating rich insights. We also facilitate the estimation of causal effects in the presence of endogeneity using machine-learning methods. Specifically, we develop novel econometric instruments that capture product differentiation (isolation) based on deeplearning models of user-generated reviews. Our empirical findings extend the current knowledge regarding the heterogeneous impact of recommender systems, reconcile contradictory prior results in the related literature, and have significant business implications.

Determining banking service attributes from online reviews: text mining and sentiment analysis

PurposeThe current study employs text mining and sentiment analysis to identify core banking service attributes and customer sentiment in online user-generated reviews. Additionally, the study explains customer satisfaction based on the identified predictors.Design/methodology/approachA total of 32,217 customer reviews were collected across 29 top banks on bankbazaar.com posted from 2014 to 2021. In total three conceptual models were developed and evaluated employing regression analysis.FindingsThe study revealed that all variables were found to be statistically significant and affect customer satisfaction in their respective models except the interest rate.Research limitations/implicationsThe study is confined to the geographical representation of its subjects' i.e. Indian customers. A cross-cultural and socioeconomic background analysis of banking customers in different countries may help to better generalize the findings.Practical implicationsThe study makes essential theoretical and managerial contributions to the existing literature on services, particularly the banking sector.Originality/valueThis paper is unique in nature that focuses on banking customer satisfaction from online reviews and ratings using text mining and sentiment analysis.

Unrestricted Attention May Not Be All You Need–Masked Attention Mechanism Focuses Better on Relevant Parts in Aspect-Based Sentiment Analysis

Aspect-Based Sentiment Analysis (ABSA) is one of the highly challenging tasks in natural language processing. It extracts fine-grained sentiment information in user-generated reviews, as it aims at predicting the polarities towards predefined aspect categories or relevant entities in free text. Previous deep learning approaches usually rely on large-scale pre-trained language models and the attention mechanism, which applies the complete computed attention weights and does not place any restriction on the attention assignment. We argue that the original attention mechanism is not the ideal configuration for ABSA, as for most of the time only a small portion of terms are strongly related to the sentiment polarity of an aspect or entity. In this paper, we propose a masked attention mechanism customized for ABSA, with two different approaches to generate the mask. The first method sets an attention weight threshold that is determined by the maximum of all weights, and keeps only attention scores above the threshold. The second selects the top words with the highest weights. Both remove the lower score parts that are assumed to be less relevant to the aspect of focus. By ignoring part of input that is claimed irrelevant, a large proportion of input noise is removed, keeping the downstream model more focused and reducing calculation cost. Experiments on the Multi-Aspect Multi-Sentiment (MAMS) and SemEval-2014 datasets show significant improvements over state-of-the-art pre-trained language models with full attention, which displays the value of the masked attention mechanism. Recent work shows that simple self-attention in Transformer quickly degenerates to a rank-1 matrix, and masked attention may be another cure for that trend.

Know Thy Context: Parsing Contextual Information from User Reviews for Recommendation Purposes

In this paper, we study an important problem of parsing contextual information from user reviews for recommendation purposes. First, we study the ways contextual information is expressed in user reviews and obtain novel insights about it. Among other things, we demonstrate that such type of information tends to appear at the beginning of the review, in longer sentences, in the sentences written in the past tense or using gerund form, and in the sentences referring to some points in time. Second, we propose a novel context parsing method for systematically extracting contextual information from user-generated reviews that rely on the insights obtained in our study. We apply the proposed method to three different Yelp applications (restaurants, hotels, and beauty &amp; spas) and demonstrate that it works well and leads to better recommendation performance than the baseline approaches. Our method systematically extracts more comprehensive sets of relevant contextual variables and corresponding phrases than the baselines. Our analysis also shows the importance of the newly discovered contextual information for recommendation purposes. The obtained results and the proposed method can help to get more comprehensive knowledge about contextual variables in a given application that leads to better recommendations.

Customer Review Provision Policies with Heterogeneous Cluster Preferences

Companies often post user-generated reviews online so that potential buyers in different clusters (age, geographic region, occupation, etc.) can learn from existing customers about the quality of an experience good and cluster preferences before purchasing. In this paper, we evaluate two common user-generated review provision policies for selling experience goods to customers in different clusters with heterogeneous preferences. The first policy is called the association-based policy (AP) under which a customer in a cluster can only observe the aggregate review (i.e., average rating) generated by users within the same cluster. The second policy is called the global-based policy (GP) under which each customer is presented with the aggregate review generated by all users across clusters. We find that, in general, the firm benefits from a policy that provides a larger number of “relevant reviews” to customers. When customers are more certain about the product quality and when clusters are more diverse, AP is more profitable than GP because it provides cluster-specific reviews to customers. Otherwise, GP is more profitable as it provides a larger number of less relevant reviews. Moreover, we propose a third provision policy that imparts the union of the information by AP and GP and show that it is more profitable for the firm. Although the third policy always renders a higher consumer welfare than GP, it may generate a lower consumer welfare than AP.This paper was accepted by Martínez-de-Albéniz Victor, operations management.