Research On Park Recommendation Algorithm Based On User Reviews

Parks are the main part of urban green space,providing a comfortable environment for people's daily activities such as sightseeing,leisure,entertainment and so on.However,due to the variety of parks,tourists need to spend more time and energy in choosing parks.Therefore,the research on park recommendation in the field of recommendation algorithm can help tourists better choose their favorite parks.Among them,the collaborative filtering algorithm is the most commonly used recommendation method in practice,but in collaborative filtering algorithm,users' historical scores are usually used to predict the unknown park rating,while ignoring their textual reviews.In order to solve this problem,based on the collaborative filtering algorithm,this paper uses emotional analys is and topic extraction to analyze user's reviews on Dianping.com,and proposes two different park recommendation algorithms based on user reviews.(1)A park recommendation algorithm based on Emotional Analysis is proposed.The algorithm first extracts the park features most concerned by users through their reviews on different parks,and then calculates the similarity of users based on their emotional preferences on various features of the park and predicts the unknown park rating for target user.Finally,the proposed algorithm is evaluated on the real park data sets,the results show that the proposed approach can significantly improve the accuracy of recommendation and effectively handle the data sparsity problem.(2)A park recommendation algorithm based on Latent Dirichlet Allocation(LDA)topic model is proposed.The proposed algorithm first uses LDA topic model to extract the statistical distribution of the park features.Secondly,it detects user preference distribution based on park features and user ratings.In order to measure the credibility of user ratings,user rating confidence level is considered to correct user preferences.Thirdly,it uses modified KL(Kullback-Leibler)divergence,i.e.JS(Jensen-Shannon)divergence to calculate the similarity between different users,and then predicts the unknown park rating for a specific user.Finally,the proposed algorithm is verified on the real park data sets and compared with other algorithms.