Research on the Price of Online Short-Term Rental Rooms Based on Fusion of User Reviews

With the rapid development of the Internet and travel market, the availability of online short-term rental rooms has emerged, making good use of limited room resources. In current research, only structured data of room characteristics, such as the location of the room, are considered in the price of the room. As unstructured textual data, online user’s reviews containing the emotional responses of users influence the price of online short-term rental rooms. In this research, user reviews are considered in a hedonic price regression model to improve the performance of regression. First, structured room characteristics are input to build a traditional hedonic price regression model. Then, fusion of emotion scores transformed from unstructured user reviews is input to build a fusion hedonic price regression model. Finally, the traditional model and the fusion model are compared statistically. Experimental results indicate that the fusion of user reviews can improve the performance of hedonic price regression model.