Search results

With the growth of social media and online communications, consumers are becoming more informed about hotels' services than ever before. They are writing online review to share their experiences, as well as reading online review before making a hotel reservation. Hotel customers considered it as reliable source and it influences customers' hotel selection. Most of these reviews reside in unstructured format, scattered across in the Internet and inherently unorganized. The purpose of this study was to use predictive text analytics to identify sentiment drivers from unstructured online reviews.

The research used sentiment classifications to analyze customers' reviews on hotels from TripAdvisor. In total, 9,286 written reviews by hotel customers were scrapped from 442 hotels in Malaysia. A detailed text analytic was conducted and was followed by a development of a theoretical framework based on the hybrid approach. AMOS was used to analyze the relationship between customer sentiments and overall review rating.

With the use of Structural Equation Modeling (SEM) and clustering technique, a list of sentiment drivers was detected, i.e. location, room, service, sleep, value for money and cleanliness. Among these variables, service quality and room facilities emerged as the most influential factors. Sentiment drivers obtained in this study provided the insights to hotel operators to improve the hotel conditions.

Although this study extended the existing literature on sentiment analysis by providing valuable insights to hoteliers, it is not without its limitations. For instance, online hotel reviews collected for this study were limited to one specific online review platform. Despite the large sample size to support and justify the findings, the generalizability power was restricted. Thus, future research should also consider and expand to other type of online review channels. Therefore, a need to examine these data reside various social media applications, i.e. Facebook, Instagram and YouTube.

This study highlights the significance of hybrid predictive model in analyzing the unstructured hotel reviews. Based on the hybrid predictive model we developed, six sentiment drivers emerged from the data analysis, i.e. location, service quality, value for money, sleep quality, room design and cleanliness. This consideration is critical due to the ever-increasing unstructured data resides in the online space. This explores the possibility of applying data analytic technique in a more efficient manner to obtain customer insights for hotel managerial consideration.

This study analyzed customer sentiments toward the hotel in Malaysia with the use of predictive text analytics technique. The main contribution was the list of sentiment drivers and the insights needed to improve the hotel conditions in Malaysia. In addition, the findings demonstrated motivating findings from different methodological perspective and provided hoteliers with the recommendation for improved review ratings.

The goal of the current research is to use different machine learning (ML) approaches to examine and predict customer reviews of food delivery apps (FDAs).

Using Google Play Scraper, data from five food delivery service providers were collected from the Google Play store. Following cleaning the reviews, the filtered texts were classified as having negative, positive, or neutral sentiments, which were then scored using two unsupervised sentiment algorithms (AFINN and Valence Aware Dictionary for sentiment Reasoning (VADER)). Furthermore, the authors employed four ML approaches to categorize each review of FDAs into the respective sentiment class.

According to the study's findings, the majority of customer reviews of FDAs were positive. This research also revealed that, while all of the methods (decision tree, linear support vector machine, random forest classifier and logistic regression) can appropriately classify the reviews into a sentiment category, support vector machines (SVM) beats the others in terms of model accuracy. The authors' study also showed that logistic regression provided the highest recall, F1 score and lowest Root Mean Square Error (RMSE) among the four ML models.

The findings aid FDAs in determining customer review behavior. The study's findings could help food apps developers better understand how customers feel about the developers' products and services. The food apps developer can learn how to use ML techniques to better understand the users' behavior.

The current study uses ML methodologies to investigate and predict consumer attitude regarding FDAs.

With the fierce competition in the cold chain logistics market, achieving and maintaining excellent customer satisfaction is the key to an enterprise's ability to stand out. This research aims to determine the factors that affect customer satisfaction in cold chain logistics, which helps cold chain logistics enterprises identify the main aspects of the problem. Further, the suggestions are provided for cold chain logistics enterprises to improve customer satisfaction.

This research uses the text mining approach, including topic modeling and sentiment analysis, to analyze the information implicit in customer-generated reviews. First, latent Dirichlet allocation (LDA) model is used to identify the topics that customers focus on. Furthermore, to explore the sentiment polarity of different topics, bi-directional long short-term memory (Bi-LSTM), a type of deep learning model, is adopted to quantify the sentiment score. Last, regression analysis is performed to identify the significant factors that affect positive, neutral and negative sentiment.

The results show that eight topics that customer focus are determined, namely, speed, price, cold chain transportation, package, quality, error handling, service staff and logistics information. Among them, speed, price, transportation and product quality significantly affect customer positive sentiment, and error handling and service staff are significant factors affecting customer neutral and negative sentiment, respectively.

The data of the customer-generated reviews in this research are in Chinese. In the future, multi-lingual research can be conducted to obtain more comprehensive insights.

Prior studies on customer satisfaction in cold chain logistics predominantly used questionnaire method, and the disadvantage of which is that interviewees may fill out the questionnaire arbitrarily, which leads to inaccurate data. For this reason, it is more scientific to discover customer satisfaction from real behavioral data. In response, customer-generated reviews that reflect true emotions are used as the data source for this research.

The purpose of this paper is to study the consumer opinion towards the low-cost airlines or low-cost carriers (LCCs) (these two terms are used interchangeably) industry in Malaysia to better understand consumers’ needs and to provide better services. Sentiment analysis is undertaken in revealing current customers’ satisfaction level towards low-cost airlines.

About 10,895 tweets (data collected for two and a half months) are analysed. Text mining techniques are used during data pre-processing and a mixture of statistical techniques are used to segment the customers’ opinion.

The results with two different sentiment algorithms show that there is more positive than negative polarity across the different algorithms. Clustering results show that both K-Means and spherical K-Means algorithms delivered similar results and the four main topics that are discussed by the consumers on Twitter are customer service, LCCs tickets promotions, flight cancellations and delays and post-booking management.

Gaining knowledge of customer sentiments as well as improvements on the four main topics discussed in this study, i.e. customer service, LCCs tickets promotions, flight cancellations or delays and post-booking management will help LCCs to attract more customers and generate more profits.

This paper provides useful insights on customers’ sentiments and opinions towards LCCs by utilizing social media information.

There is a strong prerequisite for organizations to analyze customer review behavior to evaluate the competitive business environment. The purpose of this study is to analyze and predict customer reviews of halal restaurants using machine learning (ML) approaches.

The authors collected customer review data from the Yelp website. The authors filtered the reviews of only halal restaurants from the original data set. Following cleaning, the filtered review texts were classified as positive, neutral or negative sentiments, and those sentiments were scored using the AFINN and VADER sentiment algorithms. Also, the current study applies four machine learning methods to classify each review toward halal restaurants into its sentiment class.

The experiment showed that most of the customer reviews toward halal restaurants were positive. The authors also discovered that all of the methods (decision tree, linear support vector machine, logistic regression and random forest classifier) can correctly classify the review text into sentiment class, but logistic regression outperforms the others in terms of accuracy.

The results facilitate halal restaurateurs in identifying customer review behavior.

Sentiment and emotions, according to appraisal theory, form the basis for all interactions, facilitating cognitive functions and supporting prospective customers in making sense of experiences. Emotion theory also describes human affective states that determine motives and actions. The study looks at how potential customers might react to a halal restaurant’s consensus on social media based on reviewers’ opinions of halal restaurants because emotions can be conveyed through reviews.

This study applies machine learning approaches to analyze and predict customer sentiment based on the review texts toward halal restaurants.

In this paper the authors present a study that uses Twitter to identify critical elements of customer service in the airline industry. The goal of the study was to uncover customer opinions about services by monitoring and analyzing public Twitter commentaries. The purpose of this paper is to identify elements of customer service that provide positive experiences to customers as well as to identify service processed and features that require further improvements.

The authors employed the approach of sentiment analysis as part of the netnography study. The authors processed 67,953 publicly shared tweets to identify customer sentiments about services of four airline companies. Sentiment analysis was conducted using the lexicon approach and vector-space model for assessing the polarity of Twitter posts.

By analyzing Twitter posts for their sentiment polarity the authors were able to identify areas of customer service that caused customer satisfaction, dissatisfaction as well as delight. Positive sentiments were linked mostly to online and mobile check-in services, favorable prices, and flight experiences. Negative sentiments revealed problems with usability of companies’ web sites, flight delays and lost luggage. Evidence of delightful experiences was recorded among services provided in airport lounges.

Paper demonstrates how sentiment analysis of Twitter feeds can be used in research on customer service experiences, as an alternative to Kano and SERVQUAL models.

Online hotel ratings, a form of electronic word of mouth (eWOM), are becoming increasingly important to tourism and hospitality management. Using sentiment analysis based on the big data technique, this paper aims to investigate the relationship between customer sentiment and online hotel ratings from the perspective of customers’ motives in the context of eWOM, and to further identify the moderating effects of review characteristics.

The authors first retrieve 273,457 customer-generated reviews from a well-known online travel agency in China using automated data crawlers. Next, they exploit two different sentiment analysis methods to obtain sentiment scores. Finally, empirical studies based on threshold regressions are conducted to establish the asymmetric relationship between customer sentiment and online hotel ratings.

The results suggest that the relationship between customer sentiment and online hotel ratings is asymmetric, and a negative sentiment score will exert a larger decline in online hotel ratings, compared to a positive sentiment score. Meanwhile, the reviewer level and reviews with pictures have moderating effects on the relationship between customer sentiment and online hotel ratings. Moreover, two different types of sentiment scores output by different sentiment analysis methods verify the results of this study.

The moderating effects of reviewer level and reviews with pictures offer new insights for hotel managers to make different customer service policies and for customers to select a hotel based on reviews from the online travel agency.

This paper contributes to the literature by applying big data analysis to the issues in hotel management. Based on the eWOM communication theories, this study extends previous study by providing an analysis framework for the relationship between customer sentiment and online hotel ratings from the perspective of customers’ motives in the context of eWOM.

Social media channels provide an avenue for expressing views about different services/products. However, unlike merchandise/company websites (where users can post both reviews and ratings), it is not possible to understand user's ratings for a particular service-related comment on social media unless explicitly mentioned. Predicting ratings can be beneficial for service providers and prospective customers. Additionally, predicting ratings from a user-generated content can help in developing vast data sets for recommender systems utilizing recent data. The aim of this study is to predict ratings more accurately and enhance the performance of sentiment-based predictors by combining it with the emotional content of textual data.

This study had utilized a combination of sentiment and emotion scores to predict the ratings of Twitter posts (3,509 tweets) in three different contexts, namely, online food delivery (OFD) services, online travel agencies (OTAs) and online learning (e-learning). A total of 29,551 reviews were utilized for training and testing purposes.

Results of this study indicate accuracies of 58.34%, 57.84% and 100% in cases of e-learning, OTA and OFD services, respectively. The combination of sentiment and emotion scores showed an increase in accuracies of 19.41%, 27.83% and 40.20% in cases of e-learning, OFD and OTA services, respectively.

Understanding the ratings of social media comments can help both service providers as well as prospective customers who do not spend much time reading posts but want to understand the perspectives of others about a particular service/product. Additionally, predicting ratings of social media comments will help to build databases for recommender systems in different contexts.

The uniqueness of this study is in utilizing a combination of sentiment and emotion scores to predict the ratings of tweets related to different online services, namely, e-learning OFD and OTAs.

This study aims to investigate whether and how different sentiments affect the stock market’s reaction to the American Customer Satisfaction Index (ACSI) information.

The portfolio approach, with time-varying risk factor loadings and the asset-pricing models, is borrowed from the finance literature to investigate the ACSI-performance relationship. A direct sentiment index is used to examine how investors’ optimistic, neutral and pessimistic sentiments affect the aforementioned relation.

This paper finds that customer satisfaction is a valuable intangible asset that generates positive abnormal returns. On average, investing in the Strong-ACSI Portfolio is superior to investing in the market index. Even when the stock market holds pessimistic beliefs, investors can beat the market by investing in firms that score well on customer satisfaction. The out-performance of our zero-cost, long–short ACSI strategy also confirms the mispricing of ACSI information in pessimistic periods.

Findings are limited to firms covered by the ACSI data.

Finance research has further documented evidence of the stock market under-reacting to intangible information. For example, firms with higher research and development expenditures, advertising, patent citations and employee satisfaction all earn superior returns. Literature also proves that investors efficiently react to tangible information, whereas they undervalue intangible information. In summary, combining our results and those reported in the literature, customer satisfaction is value-relevant for both investors and firm management, particularly in pessimistic periods.

This study is the first to investigate how sentiment affects the positive ACSI-performance relationship, while considering the time-varying property of risk factors. This study is also the first to show that ACSI plays a more important role during pessimistic periods. This study contributes to the growing literature on the marketing–finance interface by providing better understanding of how investor emotional states affect their perceptions and valuations of customer satisfaction.

Digital platforms have grown significantly in recent years. Although high platform failure risks (PFR) have plagued the industry, the literature has only given this issue scant treatment. Customer sentiments are crucial for platforms and have a growing body of knowledge on its analysis. However, previous studies have overlooked rich contextual information emb`edded in user-generated content (UGC). Confronting the research gap of digital platform failure and drawbacks of customer sentiment analysis, we aim to detect signals of PFR based on our advanced customer sentiment analysis approach for UGC and to illustrate how customer sentiments could predict PFR.

We develop a deep-learning based approach to improve the accuracy of customer sentiment analysis for further predicting PFR. We leverage a unique dataset of online P2P lending, i.e., a typical setting of transactional digital platforms, including 97,876 pieces of UGC for 2,467 platforms from 2011 to 2018.

Our results show that the proposed approach can improve the accuracy of measuring customer sentiment by integrating word embedding technique and bidirectional long short-term memory (Bi-LSTM). On top of that, we show that customer sentiment can improve the accuracy for predicting PFR by 10.96%. Additionally, we do not only focus on a single type of customer sentiment in a static view. We discuss how the predictive power varies across positive, neutral, negative customer sentiments, and during different time periods.

Our research results contribute to the literature stream on digital platform failure with online information processing and offer implications for digital platform risk management with advanced customer sentiment analysis.