Modeling and Sentiment Analysis of Online Reviews in Hospitality Industry

Abstract

With the great extent of use of smartphones and the internet, the online hotel booking service providers have excessively increased thus producing more user-generated content in the form of reviews and comments about the customer experience. These reviews of visited customer’s aids hotel management personnel not only to forecast the future demand but also to implement effective strategies for better service.
 It is becoming a tuff job in this scenario for the hotel management to get exact information from the wide range of reviews.
 In this analysis, it is to identify the classification of the sentiment from the customer reviews. The classification can be done with text mining approach with the source of information. Two dictionaries are developed for the usage of data classification around 431 reviews taken from Tripadvisor.com and Booking.com. Finally Latent Dirichent Allocation (LDA) modeling algorithm is applied to identify related topics and it was used to sort out the issues in consumer sentiment analysis. 
 Study findings revealed that majority of the reviews were with positive sentiments and the topics found best with hospitality domain and sentiment term were such as “food”, “hospitality”, “room”, “people”, “friendly” , “Relax”, “feelings”, and “holiday” as hospitality terms and “Strong Positive” and “Ordinary Positive” as sentiment terms.