Shared Accommodation Research On Spatial Model And Quantile Regression Model

With the prosperity and the development of the shared economy,shared accommodation has increasingly become a more popular choice for people to travel.More and more people have become the providers of shared accommodation to share their spare houses or rooms to earn extra income or social satisfaction.Then,when house-owners choose to share accommodation,what factors will be considered?And what factors play a major role on the difference of the popular shared accommodation and the poorly consumed shared accommodation,and how the influence of factors such as price on the consumption behavior of shared accommodation with different popularity changes.These questions have not been studied at present,which has become the main content of our research.To study the consumption behavior of shared accommodation,we need to know more about the consumption information of shared accommodation.Airbnb is the largest and most successful platform for shared accommodation in the world.Its official internal website provides the most comprehensive consumption information of shared accommodation in most cities of the world,including the longitude and latitude of shared accommodation,price,housing type,number of comments,etc.The consumption information of shared accommodation in Hong Kong,Berlin,and Austin provided by Airbnb platform is analyzed and studied.The three regions belong to three very important continents in the world.The three cities are prosperous cities in the world.The development of shared accommodation economy is fast in the three regions,and they have more shared accommodation sources and consumption information in the world.Mass of data can make the analysis more objective and impartial.In this paper,GeoDa software is used to make spatial distribution maps,quartile spatial distribution maps and three-dimensional spatial stereo maps using longitude and latitude information in data.It can be more vivid to see that data distribution has spatial correlation.At the same time,LMLag test,LMError test,robust LMError test and robust LMError test are obtained by ordinary regression.These tests verify the spatial measurement model hypothesis on spatial correlation,and GeoDa establishes the adjacent spatial weight matrix for modeling and analysis.Through the test results of the above four tests,the spatial lag model is established in Hong Kong and Berlin.The spatial error model is established in Austin.And the three models have passed the residuals test at the same time.The fitting effect of each model is better than that of ordinary regression.Through the spatial lag model and spatial error model,it can be seen that the consumption behavior of shared accommodation in Hong Kong and Berlin has positive spatial correlation of dependent variables,while the consumption behavior of shared accommodation in Austin has spatial correlation of error items.In order to study the main influencing factors of shared accommodation with different popularity and the influence of the same influencing factor on different popularity,this paper also establishes quantile regression models.Quantile regression can study the influence of independent variables on dependent variables at different quantiles,making the analysis more comprehensive and specific.From the results of quantile regression,we can see that housing type has a very important impact on consumers’ choices.We can see that among the most popular shared accommodation in the three regions,Hong Kong should avoid private room type.Among the moderately popular shared accommodation in the three regions,Austin should avoid private room type.In conclusion,Austin’s consumers prefer the whole rental housing type,while Berlin’s consumers prefer the shared housing type and the private housing type.Hong Kong’s consumers have no obvious preference for the shared housing type.Through the above analysis,we can give suggestions mainly around the two aspects of supply and demand to the shared accommodation platforms and the providers:1.Increase the supply of shared accommodation in the place where the spatial correlation is right.2.To satisfy the consumers’preferences,first of all,the shared accommodation platform should use the transaction information to analyze and understand consumers’ preferences in different areas,such as housing type and the limit of holding number,and provide consumers7 preferences information to the host.The shared accommodation’s landlord can adjust interior decoration according to relevant information.3.Reduce the limitations of the consumers’ choices:first of all,the shared accommodation platform needs to understand what restrictions may exist in the shared accommodation’s reservation,such as the minimum number of days to stay,gender,nationality and the inability to carry pets.For some reasons,the owners of the shared accommodations have to adopt these restrictions.The platform can provide solutions for the landlords with the same common restrictions.4.Increase interaction with consumers:the landlord who shares the residential quarters should strengthen the interaction with consumers.Through the interaction with consumers,the attraction of the shared accommodation can be better improved.At the same time,the increase of comment interaction can also attract more potential consumers.5.The determination of appropriate annual business days helps landlords reduce the general administration costs and opportunity costs.