Search results

The purpose of this study is to develop a review rating prediction method based on a supervised text mining approach for unrated customer reviews.

Using 2,851 hotel comment card (HCC) reviews, this paper manually labeled positive and negative comments with seven aspects (dining, cleanliness, service, entertainment, price, public, room) that emerged from the content of said reviews. After text preprocessing (tokenization, eliminating punctuation, stemming, etc.), two classifier models were created for predicting the reviews’ sentiments and aspects. Thus, an aggregate rating scale was generated using these two classifier models to determine overall rating values.

A new algorithm, Comment Rate (CRate), based on supervised learning, is proposed. The results are compared with another review-rating algorithm called location based social matrix factorization (LBSMF) to check the consistency of the proposed algorithm. It is seen that the proposed algorithm can predict the sentiments better than LBSMF. The performance evaluation is performed on a real data set, and the results indicate that the CRate algorithm truly predicts the overall rating with ratio 80.27%. In addition, the CRate algorithm can generate an overall rating prediction scale for hotel management to automatically analyze customer reviews and understand the sentiment thereof.

The review data were only collected from a resort hotel during a limited period. Therefore, this paper cannot explore the effect of independent variables on the dependent variable in context of larger period.

This paper provides a novel overall rating prediction technique allowing hotel management to improve their operations. With this feature, hotel management can evaluate guest feedback through HCCs more effectively and quickly. In this way, the hotel management will be able to identify those service areas that need to be developed faster and more effectively. In addition, this review rating prediction approach can be applied to customer reviews posted via online platforms for detecting review and rating reliability.

Manually analyzing textual information is time-consuming and can lead to measurement errors. Therefore, the primary contribution of this study is that although comment cards do not have rating values, the proposed CRate algorithm can predict the overall rating and understand the sentiment of the reviews in question.

The hygiene factor is always imperative when customers consider a certain restaurant, and the information contained in customer reviews can be an efficient approach to gauge a restaurant's hygiene during gaps in the official inspection. Therefore, the purpose of this study was to investigate whether information obtained from online reviews could predict the upcoming hygiene rating, specifically, evaluating the impact of both qualitative and quantitative content of reviews on the restaurant hygiene rating.

The quantitative research method with big data analytic techniques was applied in this study. In total, 127,656 pieces of data collected from 1,710 restaurants in four major cities in the USA were used in the analysis. Both quantitative factors (i.e. reviewer's numerical rating, days to review, readability, useful/funny/cool) and qualitative factors (i.e. eight emotional dimensions of textual reviews) were analyzed from the online customer reviews and considered in predicting the restaurant's hygiene rating.

Six out of eight emotional dimensions including anger, disgust, fear, sadness, surprise and trust were identified as having significant influences on the restaurant hygiene ratings. While three quantitative variables including days to review, readability and usefulness were identified with significant impacts on the dependent variable of restaurant hygiene rating.

This study opens an avenue for innovative research that establishes a connection between customers' reviews and restaurants' inspection systems. The results allow restaurants to predict an impending hygiene inspection rating upon dynamic assessment of review content and aid in adjusting hygiene measures accordingly.

Product ratings and reviews are popular tools to support buying decisions of consumers. Many e-commerce platforms now offer product ratings and reviews as ratings and reviews are valuable for online retailers. However, luxury goods industry is somewhat slow to adapt to the digital terrain. The purpose of this paper is to answer “how luxury consumers see user-generated product ratings and reviews for their online shopping experience and what important factors or values are perceived by the luxury consumers when they shop online?”

To understand how luxury consumers use product ratings and reviews before buying online, a survey with a situational set up of variations of rating, review and price options in association with a number of hypothetical luxury goods was conducted among 421 global luxury consumers out of over 6,000 people. The study was carried out from September to October 2018 for six weeks in the form of online and mobile survey. User population is high net-worth individuals or luxury consumers derived from the author’s various professional and social networks and communities. Their geographical coverage would be global, but concentrated around the major cities.

The survey shows that ratings and reviews can be important source of information for luxury consumers. Online ratings and reviews are rated as helpful by 76.01% of the participants. People who chose the highly rated one (4.8/5) over the poorly rated (3.7/5) was 86.94%, while all else such as product category, star rating and price range are about the same. Feedback from the open question survey indicates that the perceived helpfulness of rating and review systems could vary. Comparing user reviews is time-consuming because of unstructured nature of contextual reviews and the relative nature of human perception on the rating scale.

There are two aspects of ratings and reviews playing an important role for luxury consumers’ buying decision. First, it is about helpfulness of collective rating score. Luxury consumers see a user-generated rating score and use the score when they make a choice even if the rating is not an absolute, but relative figure, not exactly like the star rating system in the hotel industry. Second, there is discrepancy between the status of the brand in association with its price position and perceived value as the industry does not cope with classifying their brands in any official star rating system.

Consumers need compact and concise information about the products they need. When there are only a few potential products left in their short wish-list, full user reviews can be helpful to get more details and general opinions about the products on the short list before making a final decision. In that regard, a primary indicator that will guide through the decision-making process of the luxury consumers would be the trustworthiness of user rating of each product in an aggregated score along with a potential use of sub-ratings, which has to be visible from the product landing page.

Even if there is a wide use and ubiquitous nature of product ratings and reviews in other consumer products, the author is curious about how luxury consumers use ratings and reviews for their buying decision because there are not that many researches done previously in spite of the importance of this issue. Luxury goods industry has hit €320bn in 2017 according to Bain and Co., and 25% of the trading volume will be replaced by the digital commerce by 2025.

This paper aims to examine the effects of traditional customer satisfaction (CS) relative magnitude and social media review ratings on hotel performance and to explore which online travel intermediaries’ review ratings serve as the most reliable and valid predictor for hotel performance.

In 2014, CS and hotel performance data were collected from the internal database of full-service hotels operated and managed by a large hotel chain in the USA. Each property’s social media review ratings data were hand-collected from major online travel intermediaries and social media websites.

The results of this study indicate that social media review rating is a more significant predictor than traditional CS for explaining hotel performance metrics. Additionally, the social media review rating of TripAdvisor is the best predictor for hotel performance out of the other intermediaries.

This research contributes to the hospitality literature because it examines the incremental explanatory power of social media review rating and traditional CS on hotel performance. Among the leading online travel intermediaries, the findings show that TripAdvisor’s social media review rating has the most salient effect on hotel performance.

The result of this study provides useful practical implications for hotel marketers and revenue managers. This study assists hotel marketers and revenue managers in better allocating their budget for marketing and suggests ways for channel optimization.

The finding of this study will help revenue managers, marketing managers, and hotel owners make decisions regarding their marketing budget allocation to their social media marketing campaign and select the optimal online travel intermediaries as part of their channel management strategies.

In September 2019, Booking.com changed from the smiley-based scoring system (2.5–10) to the purely 10-point evaluation system (1–10). The smiley-based service evaluation is based on the multi-dimensional (M-D) system, whereas the purely 10-point service evaluation is based on the single-dimensional (S-D) system. This paper aims to focus on how a change in review posting policies impacts service evaluations regarding review generation and distribution.

The authors exploit the natural experiment using Booking.com when the site changed its scoring system from a multidimensional smiley-based service evaluation system to an S-D scoring system. The authors collected online reviews posted on two travel agencies (Booking.com and Priceline.com) between September 2019 and October 2020. A quasi-experimental approach, Difference-in-Differences, was used to isolate the impacts of the new scoring system from the impacts of the change in the service evaluation environment, i.e. COVID-19.

The change in the scoring system considerably alters review distributions by decreasing the portion of positive reviews but increasing the portion of highly positive reviews. Using the theory of emotion work (Hochschild, 1979, 2001), DID is also the reason that the former M-D smiley-based system could have underrated, highly positive reviews of services. Using the information transfer theory (Belkin, 1984), the authors reason the asymmetric transfer of information when users consume reviews from the older (M-D) system but are required to generate reviews on a newer (S-D) system.

The findings would provide online review platform management with a deeper understanding of the consequences of changes in service evaluations when the scoring system is changed.

Though the change in the scoring system would affect how customers evaluate the services of hotels, the causal impacts of switching to the new S-D scoring system have not yet been thoroughly covered by prior hospitality and service evaluation literature, which this research aspires to do.

In recent years, practitioners have identified a number of problems with traditional performance management (PM) systems, arguing that PM is broken and needs to be fixed. In this chapter, we review criticisms of traditional PM practices that have been mentioned by journalists and practitioners and we consider the solutions that they have presented for addressing these concerns. We then consider these problems and solutions within the context of extant scholarly research and identify (a) what organizations should do going forward to improve PM practices (i.e., focus on feedback processes, ensure accountability throughout the PM system, and align the PM system with organizational strategy) and (b) what scholars should focus research attention on (i.e., technology, strategic alignment, and peer-to-peer accountability) in order to reduce the science-practice gap in this domain.

Online reviews are often likely to be socially influenced by prior reviews. This study aims to examine key review and reviewer characteristics which may influence the social influence process.

Restaurant review data from Yelp.com are analyzed using an ordered logit model and text mining approach.

This study reveals that prior average review rating exerts a positive influence on subsequent review ratings for the same restaurant, but the effect is attenuated by the variance in existing review ratings. Moreover, social influence is stronger for consumers who had a moderate dining experience or invested less cognitive effort in writing online reviews. Compared to reviewers classified by Yelp as “elite,” non-elite reviewers appear more susceptible to the social influence of prior average review rating.

This study provides guidelines for mitigating the social influence of prior reviews and improving the accuracy of online product/service ratings, which will eventually enhance business and the reputation of online review platforms.

The findings from this study contribute to the electronic word-of-mouth (eWOM) literature and social influence literature in terms of the bidirectional nature of social influence on eWOM.

By delineating electronic word-of-mouth (e-WOM) into numerical rating and opinionated review, the purpose of this paper is to advance a research model that articulates how the provision of e-WOM can aid in alleviating consumers’ distrust of online service providers, a key determinant in the former’s adoption of the latter. The authors also endeavor to uncover the role gender plays in moderating the aforementioned relationship between e-WOM and distrust.

The research model was validated via a field survey administered on 115 college students and faculty members, who had been exposed to a custom-developed online restaurant review website. SmartPLS 2.0.M3 was employed to verify both the measurement and structural properties of the research model.

Distrust reduces male consumers’ perceptions of usefulness and ease of use toward an online service provider while increasing their adoption intention. For their female counterparts, distrust reduces both perceived ease of use and adoption intention for an online service provider. Additionally, for male consumers, only opinionated review aids in alleviating distrust. Conversely, both numerical rating and opinionated review aid in alleviating the distrust of female consumers. Moreover, in contrast to their female counterparts, male consumers are less susceptible to the influence of cognitive dissonance between numerical rating and opinionated review.

This study integrates distrust with the technology acceptance model (TAM) in an attempt to gain a deeper appreciation of technology acceptance behavior. Furthermore, this study builds on the confirmation bias theory to delineate e-WOM into numerical rating and opinionated review in order to better explicate variations in how males and females react to these two distinct forms of e-WOM. Consistent with the cognitive dissonance theory, the distinction between numerical rating and opinionated review enables further exploration of the impact of cognitive dissonance between these two forms of e-WOM on male and female consumers’ distrust of online service providers. Finally, this study unveils contrasting conflict resolution strategies adopted by male and female consumers to cope with cognitive dissonance in e-WOM.

Findings from this study yield prescriptions for practitioners in terms of how e-WOM can be harnessed to alleviate consumers’ distrust of online service provider. Whereas it is crucial for online service providers to draw on opinionated review to reduce distrust for male consumers, numerical rating should be emphasized for female consumers. This study also sensitizes practitioners to the drawback of providing both numerical rating and opinionated review at the same time due to the potential for cognitive dissonance.

This study is the first to: position distrust within the well-accepted TAM in order to enrich the understanding of technology acceptance behavior; testify to the importance of delineating between numerical rating and opinionated review due to the possibility of cognitive dissonance between these two distinct forms of e-WOM, as well as; uncover contrasting conflict resolution strategies adopted by male and female consumers to cope with cognitive dissonance in accordance with the confirmation bias theory.

This study aims to investigate how prior reviews posted by other consumers affect subsequent consumers’ evaluations and to what extent the review temporal distance can increase or reduce the social influence of prior reviews. In this study’s restaurant context, review temporal distance refers to the duration between dining time and review time of a dining experience.

The data of paired online restaurant reservations and reviews are analyzed using Ordered Logit Model. Two robustness checks are conducted to test the stability of the main estimation results.

The empirical results demonstrate that consumers’ restaurant evaluation is socially influenced by both the prior average review rating and number of prior reviews; review temporal distance has a direct negative effect on consumers’ restaurant evaluation; and review temporal distance increases the social influence of prior reviews.

This study suggests that online review matters. Both restaurants and the online review platforms should encourage consumers to share their experiences and post online reviews immediately after their consumption.

The study contributes to the literature on electronic word-of-mouth, social influence and psychological distance. First, the bi-directional nature of social influence on electronic word-of-mouth for experience-oriented product is documented. Second, for the first time, this study examines how review temporal distance could affect the social influence on consumers’ restaurant evaluation.

Online ratings (review valence) have been found to exert a strong influence on hotel room prices. This study aims to systematically synthesize research estimating the impact of online ratings on room rates using a meta-analytical method.

From major academic databases, a total of 163 estimates of the effects of online ratings on room rates were coded from 22 studies across different countries through a systematic review of relevant literature. All estimates were converted into elasticity-type effect sizes, and a hierarchical linear meta-regression was used to investigate factors explaining variations in the effect sizes.

The median elasticity of online ratings on hotel room rates was estimated to be 0.851. Meta-regression results highlighted four categories of factors moderating the size of this elasticity: data characteristics, research settings, variable measures and publication outlet. Among sub-ratings, results revealed value rating and room rating to exert the largest impact on room rates, whereas staff and cleanliness ratings demonstrated non-significant impacts.

This study provides practical implications on the relative importance of different types of online ratings for online reputation and revenue management.

This study represents the first research effort to understand factors moderating the effects of online ratings on hotel room rates based on a quantitative review of the literature. Moreover, this study provides beneficial insights into the specification of empirical hedonic pricing models and data-collection strategies, such as the selection of price variables and choices of model functional forms.