A study on prediction of user overall gratification in European continental tourism city hotels

Recommender Systems (RS) are proven to be very beneficial on e-commerce sites by providing helpful information to the customers in the decision-making process. Collaborative RS is the most popular type of these systems, and they use ratingsto find users’ opinions on specific items to determine neighborhoods between users. Traditional RS, like collaborative, content-based, knowledge-based, and hybrid systems, use two-dimensional ratings for the user and the item itself. The information about hotels in different destinations, user profiles, and their expressed reviews use to generate a new dataset. Thus, the dataset containscontextual information and the traditional two-dimensional paradigm, user, and item (hotel). The additional contextual information represents additional dimensions to make a multidimensional dataset. This article analyzes the Trip Advisor multidimensional dataset to predict the overall gratification of various hotel classes and trip types based on the importance of the contextual segment.