A Study On The Influence And Prediction Of Online Information On Shared Accommodation Bookings

With the acceleration of globalization and liberalization,more and more people are keen to experience shared life.Now,the sharing economy has become a fashionable lifestyle.After Didifs trip opened the shared transportation market,people turned their attention to shared accommodation.The biggest difference between shared accommodation and traditional accommodation is that shared accommodation reflects the communication between people,and you can live in the local people,communicate with the locals,and experience the local customs.But this will bring a problem that is the resource problem.Because the number of rooms in the local people is limited,unlike the hotel which can accommodate more people,it may bring some consumer can't book a satisfactory listing.On the other hand,since the shared accommodation is attached to the Internet platform,browsing reservations for sharing accommodation can be realized on the Internet.Therefore,the issue of sharing accommodation reservations is largely related to the information displayed on the Internet which namely online information.Based on this,this article will use online information to study the development status and pre-determined amount of shared accommodation.Based on the research at home and abroad,this paper relies on Airbnb's shared accommodation platform and collects six popular cities such as Shanghai,Beijing,Hangzhou,Dali,Chengdu and Chongqing as shown on the platform before July 7,2018.There are 1790 online listings,including 22 views,such as reviews,prices and so on,and comment data 137370.Subsequently,using descriptive statistical methods to analyze the various dimensions of online information to reflect the similarities and differences of each city.Emotion analysis is used to analyze the comment text of each listing,and the emotional value of the review text is obtained.The emotional value and online information are constructed as sample data,and the online information is fully utilized.However,most of the models currently used in the study of booking volume are regression models.Therefore,based on the predecessors,multivariate regression,ridge regression and Lasso regression in the nachine learning algorithm are used to construct the model.The results show that the model has a low degree of fit.In order to improve the fitting degree of the model,this paper uses the machine learning algorithm XGBoost to refit the data,finds that the degree of model fitting is greatly improved,and explains the influencing factors of the booking amount by using the characteristic importance of the model.This paper systematically describes the similarities and differences between the developments of shared accommodation in different cities under the sharing economy,and has a good breakthrough in booking volume prediction.It helps consumers to book their satisfactory house or help the landlord predict passenger flow volume in advance.Not only that,it can help them understand which online information will increase the possibility of booking houses.Based on the research conclusions,this article provides practical suggestions for consumers,landlords and Airbnb platforms,such as emphasizing platform display,encouraging sharing of comments,and calling on consumers to participate in the sharing process,jointly reforming the shared accommodation business and promoting shared accommodation has a better development.