Analysis of Almaty’s Restaurant Reviews through Topic Modelling

This research investigates customers’ reviews of the restaurant sector in Almaty, Kazakhstan. Although Kazakhstan offers unique foods, beverages, and establishments, few studies have investigated customers’ reviews of the restaurant business. Hence, this is a pioneering study examining problems and concerns of the restaurant sector in the eyes of customers in Almaty, Kazakhstan by implementing big data analysis. The scraped dataset from the TripAdvisor website contains more than 13,000 restaurant reviews in different languages between 2010 through 2023. Generally, English and Russian are the two dominant languages used in reviews in Almaty. Text mining techniques of topic modelling and sentiment analysis are applied in order to derive and understand the main focuses, problems, and concerns of restaurant customers. To do this the collected data between 2010–2023 is split into two roughly equal datasets that cover the periods between 2010–2017 and 2018–2023, respectively. It is revealed that while restaurant customers were less satisfied with the service process, the most positive reviews, written in both English and Russian, were obtained for the topic ‘Atmosphere and Events’. Considering the service process as a holistic process, analysing the conditions affecting good service delivery and making improvements will urge restaurant customers to form positive opinions. This study provides an opportunity for managerial, operational, and marketing departments in Almaty to improve the restaurant business in the eyes of customers. Since no precise information about the themes of the restaurants is available on the restaurant pages, no thematic distinction could be provided.