Prediction and Analysis of Hotel Ratings from Crowd-Sourced Data

Abstract

Crowdsourcing has become an essential source of information for tourists and the tourism industry. Every day, large volumes of data are exchanged among stakeholders in the form of searches, posts, shares, reviews or ratings. This paper presents a tourist-centred analysis of crowd-sourced hotel information collected from the Expedia platform. The analysis relies on Data Mining methodologies to predict trends and patterns which are relevant to tourists and businesses. First, we propose an approach to reduce the crowd-sourced data dimensionality, using correlation and Multiple Linear Regression to identify the single most representative rating. Finally, we use this rating to model the hotel customers and predict hotel ratings, using the Alternating Least Squares algorithm. In terms of contributions, this work proposes: (i) a new crowd-sourced hotel data set; (ii) a crowd-sourced rating analysis methodology; and (iii) a model for the prediction of personalised hotel ratings. © Springer International Publishing AG 2017.