Research On Sharing Accommodation Recommendation Based On Sentiment Analysis Of Reviews

The rise of sharing economy drives the emergence and development of sharing platforms in cars,bicycles,houses and other industries.Sharing accommodation refers to the short rent in the Mayi Short Rent,Airbnb and other sharing sites on the rental of individual idle accommodation.At present,the accommodation sharing platform is only in the form of e-mail to provide popular recommendations.Personalized recommendation can reduce searching time for users,so that accommodation is more likely to come to the fore,which is of great significance for users,homeowners and sharing platforms.Based on the analysis of the characteristics of the user behavior on the sharing platform,this paper proposes a personalized recommendation algorithm for the sharing accommodation,which is based on two characteristics.The first is data transaction sparse and the second is the interpersonal communication.1.Accommodation feature extraction.There are a large number of accommodation features.Only a few of features have a great impact on users.The accommodation feature is extracted from the subject by the subject model.2.Improve the latent factor model to produce a preliminary recommendation list.Sharing accommodation rises recently;the number of houses is huge,resulting in the user behavior matrix sparse problem.In this paper,we reduce the sparseness of user behavior matrix by reducing dimension,and on this basis,we improve latent factor model to generate the recommended list.3.Consider the user's affirmation to amend the initial recommendation list.The list of recommended accommodation for users,which is generated from the improved privacy semantics model,is all accommodation that they have not lived.Considering the effect of emotion on user selection,we extract emotion from user reviews and add accommodation which is highly scored to recommended list.Finally,use Airbnb open source data to experiment.The experimental results show that the proposed algorithm can effectively alleviate the sparsity of sharing accommodation recommendation,and the recommendation accuracy can be improved so that the recommended service can better meet the recommended requirements of users and homeowners.