| With the development of mobile Internet technology,the sharing economy is rapidly infiltrating on a global scale,covering areas such as transportation,accommodation,knowledge and skills,living services,medical services,and shared finance.Among them,the tourism accommodation industry is developing very rapidly,and more and more people choose sharing accommodation as their primary choice when traveling or on a business trip instead of hotel.However,because of the non-standardized nature of the sharing accommodation service,the existence of information asymmetry seriously hinders the establishment of trust between the two parties,which largely hinders the development of sharing accommodation and even the sharing economy.At the same time,the non-standardization of sharing accommodation services also determines that the quality of the host is one of the most important indicators of the quality of the listing.It is difficult for consumers to determine the quality of the host before face-to-face contact.The sharing accommodation platform provides two mechanisms to help the parties to establish trust,namely:identity authentication and review mechanism.In identity authentication,the platform developed a business badge for the host,which is an excellent host's logo,identifying those experienced hosts who provide thoughtful and attentive service.The review mechanism includes ratings and online reviews,where online reviews are textual content published after the consumer experiences the listing and are more credible.The identity authentication can make consumers understand the host(listing)to a certain extent,but it is still difficult for consumers to grasp the quality of the host.In order to get a more realistic and specific understanding of a host and the non-standardized services it provides,you can tap online reviews.Research shows that tourists like to post details of contact with the host in the comments.After reading a lot of comments,the author finds that the comments contain a large number of words expressing the host's service,which suggests that we can explore the non-standardization services of the host in the sharing accommodation by mining online reviews.In response to the above mentioned issues,this article aims to predict the sharing accommodation merchant badges.We explore business behavior through online reviews,and build a sharing accommodation merchant badge prediction model.Firstly,text mining is used to extract the behavior of the host from online reviews.A crawler based python was developed in September 2017 to crawl data for all listings in Airbnb Hangzhou,and the algorithm combines Long Short-Term Memory(LSTM)and K-mean;Then,according to the types of behaviors obtained by text mining,statistical analysis is performed on Superhosts and ordinary hosts to compare the differences in service offerings between the two types of merchants;Finally,the number and type of services for each listing are counted.Two sets of data(basic data/joining service provision data)are used to compare and analyze the badge prediction performance of different algorithms,and the optimal model tuning parameters are selected to build sharing accommodation merchant badge prediction model.The results show that there are six main types of services provided by the landlord:replies or communications,fruits,drinks or snacks,three meals or night snacks,free shuttle or helping with luggage and travel guides,and chats;The service offering has a lot to do with the host badge.The Superhosts offers more services than non-Superhosts,and consumers tend to post higher ratings on listings with the "Superhost" badge.On the same model,the prediction accuracy of data 2 is higher than data 1,indicating that the host's services extracted by the article provides effective information support for the merchant badge prediction,which provides suggestions for the platform to develop a more complete merchant badge evaluation rule,and guides the direction of the operation of platform merchants. |
PDF Full Text Request