Research On Recommendation System Based On User Trust Of Shared Economic Platform

Since the sharing economy was first written into the government report in 2016,it has been actively supported in the national level,especially in the field of shared housing,the effective use of idle resources has improved people's travel accommodation to some extent.However,inevitably,the issue of shared trust has surfaced.This paper studies and analyzes the predecessors' research on the sharing economy and the trust of e-commerce,take the "xiaozhuduanzu" platform,the largest short-term rental platform in China as an example,studies the personal characteristics of the landlord on it and the influence of the tenant's commentary information on shared trust.I have climbed the basic personal information of 31 active landlords in Beijing,and the corresponding guest comment data.Combined with the characteristics of traditional e-commerce shared by the economic platform,based on the relationship between trust and user interaction behavior,The LSTM long short-term memory network fully exploits the user's commentary information,obtains the emotional score as a measure of trust,and based on the tenant-host trust matrix and the landlord's personal character matrix,builds a recommendation system using User-Specific Feature-Based Similarity Models(UFSM)for Top-N Recommendation.That provides a personalized recommendation for the tenant to select the landlord.This research started from a new perspective,fully exploited the existing platform data,and applied the theoretical research results of the trust problem to the recommendation system,and made recommendations for the transaction selection of users sharing the short-term rental platform to enhance the user experience;It also provides new ideas for research in the field of shared economy.