Detecting fake-review buyers using network structure: Direct evidence from Amazon.

Online reviews significantly impact consumers' decision-making process and firms' economic outcomes and are widely seen as crucial to the success of online markets. Firms, therefore, have a strong incentive to manipulate ratings using fake reviews. This presents a problem that academic researchers have tried to solve for over two decades and on which platforms expend a large amount of resources. Nevertheless, the prevalence of fake reviews is arguably higher than ever. To combat this, we collect a dataset of reviews for thousands of Amazon products and develop a general and highly accurate method for detecting fake reviews. A unique difference between previous datasets and ours is that we directly observe which sellers buy fake reviews. Thus, while prior research has trained models using laboratory-generated reviews or proxies for fake reviews, we are able to train a model using actual fake reviews. We show that products that buy fake reviews are highly clustered in the product reviewer network. Therefore, features constructed from this network are highly predictive of which products buy fake reviews. We show that our network-based approach is also successful at detecting fake review buyers even without ground truth data, as unsupervised clustering methods can accurately identify fake review buyers by identifying clusters of products that are closely connected in the network. While text or metadata can be manipulated to evade detection, network-based features are more costly to manipulate because these features result directly from the inherent limitations of buying reviews from online review marketplaces, making our detection approach more robust to manipulation.

How should medical society face patient feedback in online review platforms?

In the medical field, the importance of online reviews is escalating. However, the complexity of responding to these reviews is profound, as such anonymous critiques may encompass not only emotionally distressing content but also potentially malicious criticisms directed at healthcare professionals. While recognizing the vital role of patient feedback, there exists a necessity for a collective approach to managing online commentary. This effort seeks to strike a balance between patient satisfaction and the safeguarding of healthcare practitioners and administrative staff. We believe the global medical community must establish guidelines to effectively handle such scenarios, thereby contributing to the sustainability of patient-centered services.

Examining the Role of Physician Characteristics in Web-Based Verified Primary Care Physician Reviews: Observational Study.

BACKGROUND: Doctor review websites have become increasingly popular as a source of information for patients looking to select a primary care provider. Zocdoc is one such platform that allows patients to not only rate and review their experiences with doctors but also directly schedule appointments. This study examines how several physician characteristics including gender, age, race, languages spoken in a physician's office, education, and facial attractiveness impact the average numerical rating of primary care doctors on Zocdoc. OBJECTIVE: The aim of this study was to investigate the association between physician characteristics and patient satisfaction ratings on Zocdoc. METHODS: A data set of 1455 primary care doctor profiles across 30 cities was scraped from Zocdoc. The profiles contained information on the physician's gender, education, and languages spoken in their office. Age, facial attractiveness, and race were imputed from profile pictures using commercial facial analysis software. Each doctor profile listed an average overall satisfaction rating, bedside manner rating, and wait time rating from verified patients. Descriptive statistics, the Wilcoxon rank sum test, and multivariate logistic regression were used to analyze the data. RESULTS: The average overall rating on Zocdoc was highly positive, with older age, lower facial attractiveness, foreign degrees, allopathic degrees, and speaking more languages negatively associated with the average rating. However, the effect sizes of these factors were relatively small. For example, graduates of Latin American medical schools had a mean overall rating of 4.63 compared to a 4.77 rating for US graduates (P<.001), a difference roughly equivalent to a 2.8% decrease in appointments. On multivariate analysis, being Asian and having a doctor of osteopathic medicine degree were positively associated with higher overall ratings, while attending a South Asian medical school and speaking more European and Middle Eastern languages in the office were negatively associated with higher overall ratings. CONCLUSIONS: Overall, the findings suggest that age, facial attractiveness, education, and multilingualism do have some impact on web-based doctor reviews, but the numerical effect is small. Notably, bias may play out in many forms. For example, a physician's appearance or accent may impact a patient's trust, confidence, or satisfaction with their physician, which could in turn influence their take-up of preventative services and lead to either better or worse health outcomes. The study highlights the need for further research in how physician characteristics influence patient ratings of care.

Social media use and patient ratings in shoulder and elbow surgeons: how many "likes" for five stars?

BACKGROUND: Social media plays an important role in healthcare and physician selection by facilitating direct communication with patients and impacting physician ratings. A concern however is the increased online scrutiny and negative impact on patient satisfaction with these connections. This study aimed to investigate whether social media activity by fellowship-trained shoulder and elbow surgeons impacts patient's perceptions and ratings on physician review websites (PRWs). METHODS: The American Shoulder and Elbow Surgeons physician directory was used to identify currently practicing shoulder and elbow surgeons in the United States. Physician ratings were collected from Healthgrades, Google reviews, and Vitals. The surgeons were divided into two groups: social media users (SMU) and non- SMU (NSMU). The association of social media use with online physician ratings was evaluated using simple and multilinear regressions. RESULTS: A total of 385 American Shoulder and Elbow Surgeons surgeons were included and 21.3% were SMU. SMU were younger (mean age, 48 years) compared to NSMU (mean age, 51 years) (P = .01), all other demographics were comparable including sex (P = .797), medical degree (P = .114), and geographic location within the United States (P = .49). SMU had significantly higher ratings on Healthgrades (P < .001) and Vitals (P < .001). However, social media use did not impact the total number of ratings on PRWs. Additionally, surgeons who utilized Facebook had higher physician ratings and number of website reviews on Healthgrades (P = .028 and P = .014, respectively). In addition, surgeons who used Twitter had higher ratings on Healthgrades (P < .001) and Vitals (P = .001). Surgeons with a greater average number of likes per post on Twitter had significantly higher average ratings across all three sites (P = .004). Surgeons with a greater number of Twitter followers and greater average number of likes per post had significantly higher average ratings on Healthgrades (P = .052 and P = .005, respectively) while surgeons with a greater average number of likes per post had significantly higher average ratings on Vitals (P = .006). Finally, surgeons with a greater average posting frequency on Instagram had significantly higher average ratings across all websites (P = .029). CONCLUSION: Shoulder and elbow surgeons who use Twitter and Facebook had significantly higher online ratings. However, the increased use of these platforms in terms of postcontent, postfrequency, comments, and number of followers was not as influential on PRWs. This suggests that social media is an important marketing and outreach method for orthopedic surgeons to improve their ratings and patient reviews.

Safety Concerns in Mobility-Assistive Products for Older Adults: Content Analysis of Online Reviews.

BACKGROUND: Older adults who have difficulty moving around are commonly advised to adopt mobility-assistive devices to prevent injuries. However, limited evidence exists on the safety of these devices. Existing data sources such as the National Electronic Injury Surveillance System tend to focus on injury description rather than the underlying context, thus providing little to no actionable information regarding the safety of these devices. Although online reviews are often used by consumers to assess the safety of products, prior studies have not explored consumer-reported injuries and safety concerns within online reviews of mobility-assistive devices. OBJECTIVE: This study aimed to investigate injury types and contexts stemming from the use of mobility-assistive devices, as reported by older adults or their caregivers in online reviews. It not only identified injury severities and mobility-assistive device failure pathways but also shed light on the development of safety information and protocols for these products. METHODS: Reviews concerning assistive devices were extracted from the "assistive aid" categories, which are typically intended for older adult use, on Amazon's US website. The extracted reviews were filtered so that only those pertaining to mobility-assistive devices (canes, gait or transfer belts, ramps, walkers or rollators, and wheelchairs or transport chairs) were retained. We conducted large-scale content analysis of these 48,886 retained reviews by coding them according to injury type (no injury, potential future injury, minor injury, and major injury) and injury pathway (device critical component breakage or decoupling; unintended movement; instability; poor, uneven surface handling; and trip hazards). Coding efforts were carried out across 2 separate phases in which the team manually verified all instances coded as minor injury, major injury, or potential future injury and established interrater reliability to validate coding efforts. RESULTS: The content analysis provided a better understanding of the contexts and conditions leading to user injury, as well as the severity of injuries associated with these mobility-assistive devices. Injury pathways-device critical component failures; unintended device movement; poor, uneven surface handling; instability; and trip hazards-were identified for 5 product types (canes, gait and transfer belts, ramps, walkers and rollators, and wheelchairs and transport chairs). Outcomes were normalized per 10,000 posting counts (online reviews) mentioning minor injury, major injury, or potential future injury by product category. Overall, per 10,000 reviews, 240 (2.4%) described mobility-assistive equipment-related user injuries, whereas 2318 (23.18%) revealed potential future injuries. CONCLUSIONS: This study highlights mobility-assistive device injury contexts and severities, suggesting that consumers who posted online reviews attribute most serious injuries to a defective item, rather than user misuse. It implies that many mobility-assistive device injuries may be preventable through patient and caregiver education on how to evaluate new and existing equipment for risk of potential future injury.

Doctors ranking through heterogeneous information: The new score functions considering patients' emotional intensity.

With the popularity of the Internet and the growing complexity of COVID-19, more and more patients tend to consult doctors online. With the difficulty of doctor selection caused by a massive amount of information, this study proposes a hybrid multi-criteria decision-making framework, which can model patients' emotional intensity through heterogeneous information and rank doctors. Firstly, online reviews (ORs) are transformed into probabilistic linguistic term sets through sentiment analysis. Then, new score functions are proposed considering the nonlinear influence of doctors' information and the patients' negative bias toward ORs. Next, a method of weight determination combining the Term Frequency Inverse Document Frequency and the Decision-making Trial and Evaluation Laboratory method is proposed. Finally, the proposed score functions are applied to the Combined Compromise Solution (CoCoSo) method to aggregate information and rank doctors. The proposed method is verified in a case study on haodf.com. The results show that considering the emotional intensity of heterogeneous information will make the recommendations more realistic. Comparative analysis and sensitivity analysis are further performed to illustrate the availability and effectiveness of the proposed method.

A Combined Weighting Based Large Scale Group Decision Making Framework for MOOC Group Recommendation.

Massive open online courses (MOOC) are free learning courses based on online platforms for higher education, which not only promote the open sharing of learning resources, but also lead to serious information overload. However, there are many courses on MOOCs, and it can be difficult for users to choose courses that match their individual or group preferences. Therefore, a combined weighting based large-scale group decision-making approach is proposed to implement MOOC group recommendations. First, based on the MOOC operation mode, we decompose the course content into three stages, namely pre-class, in-class, and post-class, and then the curriculum-arrangement-movement- performance evaluation framework is constructed. Second, the probabilistic linguistic criteria importance through intercriteria correlation method is employed to obtain the objective weighting of the criterion. Meanwhile, the word embedding model is utilized to vectorize online reviews, and the subjective weighting of the criteria are acquired by calculating the text similarity. The combined weighting then can be obtained by fusing the subjective and objective weighting. Based on this, the PL-MULTIMIIRA approach and Borda rule is employed to rank the alternatives for group recommendation, and an easy-to-use formula for group satisfaction is proposed to evaluate the effect of the proposed method. Furthermore, a case study is conducted to group recommendations for statistical MOOCs. Finally, the robustness and effectiveness of the proposed approach were verified through sensitivity analysis as well as comparative analysis.

A sentiment analysis driven method based on public and personal preferences with correlated attributes to select online doctors.

This paper proposes a method to assist patients in finding the most appropriate doctor for online medical consultation. To do that, it constructs an online doctor selection decision-making method that considers the correlation attributes, in which the measure of attribute correlation is derived from the history real decision data. To combine public and personal preference with correlated attributes, it proposes a Choquet integral based comprehensive online doctor ranking method. In detail, a two stage classification model based on BERT (Bidirectional Encoder Representations from Transformers) is used to extract service features from unstructured text reviews. Then, 2-additive fuzzy measure is adopted to represent the patient public group aggregated attribute preference. Next, a novel optimization model is proposed to combine the public preference and personal preference. Finally, a case study of dxy.com is carried out to illustrate the procedure of the method. The comparison result between proposed method and other traditional MADM (multi-attribute decision-making) methods prove its rationality.

Text Mining Approaches for Postmarket Food Safety Surveillance Using Online Media.

Food contamination and food poisoning pose enormous risks to consumers across the world. As discussions of consumer experiences have spread through online media, we propose the use of text mining to rapidly screen online media for mentions of food safety hazards. We compile a large data set of labeled consumer posts spanning two major websites. Utilizing text mining and supervised machine learning, we identify unique words and phrases in online posts that identify consumers' interactions with hazardous food products. We compare our methods to traditional sentiment-based text mining. We assess performance in a high-volume setting, utilizing a data set of over 4 million online reviews. Our methods were 77-90% accurate in top-ranking reviews, while sentiment analysis was just 11-26% accurate. Moreover, we aggregate review-level results to make product-level risk assessments. A panel of 21 food safety experts assessed our model's hazard-flagged products to exhibit substantially higher risk than baseline products. We suggest the use of these tools to profile food items and assess risk, building a postmarket decision support system to identify hazardous food products. Our research contributes to the literature and practice by providing practical and inexpensive means for rapidly monitoring food safety in real time.

Hotel guest satisfaction during COVID-19 outbreak: The moderating role of crisis response strategy.

Taking appropriate strategies in response to the COVID-19 crisis has presented significant challenges to the hospitality industry. Based on situational crisis communication theory (SCCT), this study aims to examine how the hotel industry has adopted strategies in shaping customers' experience and satisfaction. A mixed-method approach was employed by analysing 6556 COVID-19 related online reviews. The qualitative findings suggest that 'rebuild strategies' dominated most hotels' response to the COVID-19 crisis while the quantitative findings confirm the direct impact of affective evaluation and cognitive effort on customer satisfaction. The results further reveal that hotels' crisis response strategies moderate the effects of affective evaluation and cognitive effort on customer satisfaction. The study contributes to new knowledge on health-related crisis management and expands the application of SCCT in tourism research.

The interpersonal effects of distinct emotions in online reviews.

Emotional expressions in online reviews affect reviews' informative value. By comparing high and low arousal emotions with a negative and positive valence, the current research demonstrates that the effects of emotional expressions in online reviews are determined not by the level of arousal, but by the perceived rationality of the reviewer and the perceived appropriateness of the emotional expression. In a lab experiment (N = 242) among university students, and an online experiment (N = 252) on Prolific Academic involving native English speakers, participants read an online restaurant review with the negative emotions anger, disappointment, or disgust, or with the positive emotions happiness, excitement, or contentment. Results showed that readers of online reviews considered expressions of anger more inappropriate than expressions of disappointment or disgust; this led them to judge the reviewer as more irrational, which decreased the informative value of the review. As a consequence, angry reviews led to less negative restaurant evaluations and stronger intentions to visit the restaurant than reviews expressing disappointment or disgust. We found no differences between contentment and happiness (Study 1), or between contentment and excitement (Study 2). Our findings underscore the importance of studying the effects of discrete emotions in online reviews.

Hearing aid acquisition and ownership: what can we learn from online consumer reviews?

OBJECTIVE: To explore the publicised opinions of consumers actively participating in online hearing aid reviews. DESIGN: A retrospective design examining data generated from an online consumer review website (www.HearingTracker.com). Qualitative data (open text responses) were analysed using the open source automated topic modelling software IRaMuTeQ (http://www.iramuteq.org/) to identify themes. Outputs were compared with quantitative data from the consumer reviews (short response questions exploring hearing aid performance and benefit, and some meta-data such as hearing aid brand and years of hearing aid ownership). STUDY SAMPLE: 1378 online consumer hearing aid reviews. RESULTS: Six clusters within two domains were identified. The domain Device Acquisition included three clusters: Finding the right provider, device and price-point; Selecting a hearing aid to suit the hearing loss; Attaining physical fit and device management skills. The domain Device Use included three clusters: Smartphone streaming to hearing aids; Hearing aid adjustment using smartphone; and Hearing in noise. CONCLUSIONS: Although online hearing aid consumers indicate positive performance on multiple-choice questions relating to hearing aid performance and benefit, their online reviews describe a number of barriers limiting their success. Hearing healthcare clinicians must employ a personalised approach to audiological rehabilitation to ensure individual clients' needs are met.

Social media use by shoulder and elbow surgeons increases the number of ratings on physician review websites.

BACKGROUND: The use of physician review websites by patients is increasingly frequent. One potential way for shoulder and elbow surgeons to enhance their online reputation and attract patients is via social media, yet its impact is unknown. This study sought to analyze the effect of social media use on the number of online ratings and overall rating of shoulder and elbow surgeons. We secondarily studied variables affecting social media use. METHODS: The American Shoulder and Elbow Surgeons directory was probed to identify practicing surgeons. Surgeon evaluation data, including ratings, comments, and reviews, were compiled from 3 physician review websites (Google, Healthgrades, and Vitals). Google was queried to assess for a professional Facebook, Twitter, or Instagram account, as well as obtain information on surgeon training, practice location, and other demographic variables. RESULTS: A total of 646 surgeons met the inclusion criteria (93.8% male and 6.2% female surgeons). Overall, 37% had a professional social media account (Facebook, 23.1%; Twitter, 24.1%; and Instagram, 9.4%). Linear regression analysis showed that Facebook use correlated with an average increase of 48.6 in the number of ratings. No social media platform correlated with physician rating. The surgeons more likely to use social media were those who graduated residency in 2000 or later (40.8% vs. 29.2%), those who practiced in cities with higher populations (mean, 1188.9 vs. 708.4 [per 1000]), and those with more surgeons practicing in the same city (mean, 7.0 vs. 5.0). CONCLUSION: The majority of shoulder and elbow surgeons do not have a professional social media account. Those using this platform are younger and located in more populous cities with more competition. Having a professional social media profile was not correlated with ratings, but there was a positive association with the number of online ratings, and Facebook had the strongest association. Surgeon ratings are overwhelmingly positive with minimal variability; therefore, a high number of ratings confirms a surgeon's exceptional reputation. The residency graduation year, city population, and number of nearby surgeons affected ratings, although the absolute differences were minimal. For shoulder and elbow surgeons, a professional social media account correlates with an increase in the number of online physician ratings. Recent graduates practicing in competitive locations may feel increased pressure to leverage this in an attempt to build their practices.

Getting to the Bottom of Patient Experiences with Genital Gender Affirmation Surgery: A Qualitative Analysis of Online Reviews.

BACKGROUND: Patients undergoing gender-affirming surgery seek information from online sources including online reviews written by peers. We aimed to conduct a qualitative analysis of the information discussed in online reviews related to genital gender-affirming surgery and evaluate the topics driving positive/negative reviews. METHODS: Reviews for genital gender-affirming surgery (vaginoplasty, metoidioplasty, and phalloplasty) were identified on three popular review platforms: Google, Yelp, and RealSelf. Content was analyzed line by line using a conventional inductive content analysis to identify recurring themes. Individual statements were marked as either having a positive or negative sentiment. Median rating was calculated and compared across platforms (max score 5). Associations between theme/subthemes and sentiment were also analyzed using Chi-squared test. RESULTS: A total of 129 reviews were analyzed and 433 codes were identified, the majority of which were positive (n=372; 85.9%). Three overarching themes described factors important to patient experience: surgeon medical, surgeon non-medical, and non-surgeon; with surgeon medical being the most popular. Fifteen subthemes comprised these themes, the most popular being interactions with supporting staff, surgeon bedside manner, and overall postoperative result. There was no difference in median review ratings between platforms (5 across all platforms; p=0.452). There was no association between sentiment and themes or subthemes (p=0.187 and p=0.578, respectively). CONCLUSIONS: This study is the first to analyze online reviews of genital gender-affirming surgery. The majority of patients gave positive ratings and the qualitative content had mostly positive sentiment. Salient themes not only include surgeon medical care and outcome, but other nonsurgical elements that formulate the patient's experience as a whole. LEVEL OF EVIDENCE V: This journal requires that authors assign a level of evidence to each article. For a full description of these Evidence-Based Medicine ratings, please refer to the Table of Contents or the online Instructions to Authors www.springer.com/00266 .

The impact of online reviews in the presence of customer returns.

We develop a duopoly model to examine how online reviews influence the decisions of two competing online sellers who sell products of differentiated quality under different returns policies. We derive the competing sellers' optimal decisions on price and returns policy with and without online reviews, and we find that online reviews have greater impact on the high-quality seller than on the low-quality seller. If the salvage value of the product is relatively low, the seller has less opportunity to benefit from online reviews when it offers an MBG, as compared to a no-refund policy. The impact of online reviews on the competition between the two sellers has a "symmetric effect area," where reviews may either weaken or intensify the price competition between the two sellers when they both offer a no-refund policy, but always intensify the competition if they both offer an MBG. We have identified the conditions under which online reviews lead to a win-win, or benefit one seller, or present a prisoner's dilemma for the two online sellers. We also show that MBGs at both sellers help mitigate the prisoner's dilemma if the net salvage value at both sellers is sufficiently high.

Comparative study of deep learning models for analyzing online restaurant reviews in the era of the COVID-19 pandemic.

Online reviews remain important during the COVID-19 pandemic as they help customers make safe dining decisions. To help restaurants better understand customers' needs and sustain their business under current circumstance, this study extracts restaurant features that are cared for by customers in current circumstance. This study also introduces deep learning methods to examine customers' opinions about restaurant features and to detect reviews with mismatched ratings. By analyzing 112,412 restaurant reviews posted during January-June 2020 on Yelp.com, four frequently mentioned restaurant features (e.g., service, food, place, and experience) along with their associated sentiment scores were identified. Findings also show that deep learning algorithms (i.e., Bidirectional LSTM and Simple Embedding + Average Pooling) outperform traditional machine learning algorithms in sentiment classification and review rating prediction. This study strengthens the extant literature by empirically analyzing restaurant reviews posted during the COVID-19 pandemic and discovering suitable deep learning algorithms for different text mining tasks.

"His lack of a mask ruined everything." Restaurant customer satisfaction during the COVID-19 outbreak: An analysis of Yelp review texts and star-ratings.

The aim of the study was to provide practical advice to restaurant managers for improving star ratings as well as information for researchers on how the pandemic has impacted established determinants of satisfaction. The study examined criteria used by restaurant customers in assigning star-ratings on Yelp during the COVID-19 pandemic using keyword analysis and Multiple Correspondence Analysis. In evaluating restaurants, the reviewers focused on service, overall experience, and food quality. Service was discussed in relation to the pandemic and included safety of the dine-in experience, contrasted with take-out options and compliance with COVID-19 guidelines. These criteria applied differently with lower-star reviews focusing on safety, social distancing, and mask policies. Higher-star reviews focused on take-out/delivery services, high-quality food, and an overall positive experience. The study provides valuable contributions to our understanding of how the COVID-19 pandemic will impact the restaurant sector in a post-pandemic world.

Exploring Types of Information Sources Used When Choosing Doctors: Observational Study in an Online Health Care Community.

BACKGROUND: Patients attempt to make appropriate decisions based on their own knowledge when choosing a doctor. In this process, the first question usually faced is that of how to obtain useful and relevant information. This study investigated the types of information sources that are used widely by patients in choosing a doctor and identified ways in which the preferred sources differ in various situations. OBJECTIVE: This study aims to address the following questions: (1) What is the proportion in which each of the various information sources is used? (2) How does the information source preferred by patients in choosing a doctor change when there is a difference in the difficulty of medical decision making, in the level of the hospital, or in a rural versus urban situation? (3) How do information sources used by patients differ when they choose doctors with different specialties? METHODS: This study overcomes a major limitation in the use of the survey technique by employing data from the Good Doctor website, which is now China's leading online health care community, data which are objective and can be obtained relatively easily and frequently. Multinomial logistic regression models were applied to examine whether the proportion of use of these information sources changes in different situations. We then used visual analysis to explore the question of which type of information source patients prefer to use when they seek medical assistance from doctors with different specialties. RESULTS: The 3 main information sources were online reviews (OR), family and friend recommendations (FR), and doctor recommendations (DR), with proportions of use of 32.93% (559,345/1,698,666), 23.68% (402,322/1,698,666), and 17.48% (296,912/1,698,666), respectively. Difficulty in medical decision making, the hospital level, and rural-urban differences were significantly associated with patients' preferred information sources for choosing doctors. Further, the sources of information that patients prefer to use were found to vary when they looked for doctors with different medical specialties. CONCLUSIONS: Patients are less likely to use online reviews when medical decisions are more difficult or when the provider is not a tertiary hospital, the former situation leading to a greater use of online reviews and the latter to a greater use of family and friend recommendations. In addition, patients in large cities are more likely to use information from online reviews than family and friend recommendations. Among different medical specialties, for those in which personal privacy is a concern, online reviews are the most common source. For those related to children, patients are more likely to refer to family and friend recommendations, and for those related to surgery, they value doctor recommendations more highly. Our results can not only contribute to aiding government efforts to further promote the dissemination of health care information but may also help health care industry managers develop better marketing strategies.

Assessing Patient Experience and Healthcare Quality of Dental Care Using Patient Online Reviews in the United States: Mixed Methods Study.

BACKGROUND: Over the last two decades, patient review websites have emerged as an essential online platform for doctor ratings and reviews. Recent studies suggested the significance of such websites as a data source for patients to choose doctors for healthcare providers to learn and improve from patient feedback and to foster a culture of trust and transparency between patients and healthcare providers. However, as compared to other medical specialties, studies of online patient reviews that focus on dentists in the United States remain absent. OBJECTIVE: This study sought to understand to what extent online patient reviews can provide performance feedbacks that reflect dental care quality and patient experience. METHODS: Using mixed informatics methods incorporating statistics, natural language processing, and domain expert evaluation, we analyzed the online patient reviews of 204,751 dentists extracted from HealthGrades with two specific aims. First, we examined the associations between patient ratings and a variety of dentist characteristics. Second, we identified topics from patient reviews that can be mapped to the national assessment of dental patient experience measured by the Patient Experience Measures from the Consumer Assessment of Healthcare Providers and Systems (CAHPS) Dental Plan Survey. RESULTS: Higher ratings were associated with female dentists (t71881=2.45, P<.01, g=0.01), dentists at a younger age (F7, 107128=246.97, P<.001, g=0.11), and those whose patients experienced a short wait time (F4, 150055=10417.77, P<0.001, g=0.18). We also identified several topics that corresponded to CAHPS measures, including discomfort (eg, painful/painless root canal or deep cleaning), and ethics (eg, high-pressure sales, and unnecessary dental work). CONCLUSIONS: These findings suggest that online patient reviews could be used as a data source for understanding the patient experience and healthcare quality in dentistry.

The 7 Ps marketing mix of home-sharing services: Mining travelers' online reviews on Airbnb.

The 7 Ps model is a very useful tool in helping service firms solve managerial issues in marketing. Guided by the 7 Ps marketing mix framework, a big-data, supervised machine learning analysis was performed with 1,148,062 English reviews of 37,092 Airbnb listings in San Francisco and New York City. The results disclose similar patterns in both markets, where travelers shared their experience about Service Product and Physical Evidence most often; Price and Promotion were the least mentioned elements. Furthermore, through a series of comparisons of Airbnb's 7 Ps marketing mix among the listings managed by different types of hosts, multi-unit and single-unit hosts seem to offer similar services with a small observable difference; whereas superhosts and the ordinary hosts deliver different services. This study makes valuable methodological contributions and provides practical marketing insights for hoteliers and the hosts and webmasters on home-sharing websites. Policymakers should pay special attention to multi-unit hosts.