NEHA October 2022 Journal of Environmental Health

October 2022 • Journal of Environmental Health 17 3. cross-contamination and allergens, 4. time and temperature control, and 5. cleaning and sanitation (National Restaurant Association Educational Foundation, 2022). Food handlers receive this training to enhance their food safety knowledge and apply these principles toward reducing the risks of physical, chemical, and biological contaminants in food. To ensure that food service establishments adhere to food safety principles, public health o€cials at city and county levels inspect food service establishments and issue violations for poor practices. An example of a violation would be indicators of pest presence in establishments. Pests can be vectors of pathogenic bacteria; flies and cockroaches are two of the most prevalent pests in restaurant settings (Morrison, 2007). A 2006 study in restaurants discovered cockroaches that carried Salmonella spp., E. coli O157:H7, Shigella spp., Staphylococcus aureus, andBacillus cereus—all critical foodborne pathogens (Tachbele et al., 2006). Similarly, several studies have found that common houseflies can be vectors of Shigella spp. andSalmonella spp. (Poravi et al., 2014). Additionally, health inspectors look for conditions in food service establishments that can denote bacterial activity. For example, biofilms are matrices of bacterial cells and sugars that form sticky substances on drains and in machinery and thus can clog systems and harbor infectious bacteria (Mair-Jenkins et al., 2017). Inspectors check for food preparation and service surfaces that have not been properly cleaned or sanitized. While restaurant inspections and sta— trainings are crucial in preventing FBIs, restaurants still account for 61% of FBI outbreaks in the U.S. (DeweyMattia et al., 2018). Therefore, innovative methods of monitoring and preventing FBI are necessary to decrease the annual number of outbreaks nationwide. Yelp as an Online Customer Food Safety Platform Yelp is an industry-leading crowdsourced review forum where restaurant customers (i.e., Yelpers) post reviews of businesses and rate their satisfaction as a customer on a scale of 1 to 5 stars. Yelpers primarily report on the meal quality, service level, and restaurant ambience. Customers are free, however, to write whatever they want in their online reports. Yelp aggregates the rating from each individual review for a specific restaurant into an average review rating for the restaurant. In recent years, Yelp has evolved into a tool that researchers can use to identify isolated FBI events. A 2015 study determined FBI-related reviews on Yelpwere often “extremely detailed,” even mentioning specific foods that had been implicated in foodborne outbreak reports by CDC (Nsoesie, et al., 2014). Another study identified multiple unreported FBI outbreaks in New York City via Yelp reviews (Harrison et al., 2014). As another example, the St. Louis Department of Health implemented a webbased dashboard and captured relevant tweets that reported an FBI outbreak that resulted in more filed reports than previously reported in St. Louis, Missouri (Harris et al., 2017). The aforementioned studies demonstrate a discrepancy between what individuals report to their local public health department and what restaurant customers post on online review platforms. A 2016 study that focused on discrepancies between Yelp ratings and health inspection scores found that while a larger volume of Yelp reviews for one restaurant typically was correlated with higher Yelp ratings, there was no significant relationship between health inspection scores and Yelp ratings (Park et al., 2016). Business Analytics in Food Service Business analytics refers to the methods and techniques that are used to evaluate the performance of a company (Liebowitz, 2011). An example of a business analytics technique is text mining, which is the process of knowledge discovery from textual databases and extraction of significant patterns from unstructured text documents (Tan, 2000). The process of text mining begins with the creation of a dictionary of keywords that the software will then perform a search on. In the context of online reviews, text mining can be used to measure customers’ emotional expressions (Lee et al., 2017). Thus, text mining customer-generated reviews can improve the understanding of customer behavior and trends, which can drive businesses to implement strategic change (Chau & Xu, 2012). Prior studies have empirically and conceptually examined the factors a—ecting customer overall satisfaction ratings online. Business Analytics Methodology Data Pre-Processing STEP 1 STEP 2 STEP 3 STEP 4 Text Pre-Processing Data and Text Analysis Business Analytics Online Restaurant Reviews Dictionary Development Extraction and Cleaning 231,381 Reviews From 954 Restaurants in Houston, Texas Keyword Frequency in CustomerGenerated Reviews Text Analysis With XLMiner and WordStat 8.0.8 Statistical and Descriptive Analysis With SPSS and Excel Results Results Data and Text Preparation Data and Text Collection Relationship Between Food Safety and Customer Satisfaction Context and Negation Filtration Frequency of Keywords FIGURE 1

RkJQdWJsaXNoZXIy NTU5MTM=