Research And Application Of Recommendation Algorithm Based On Home Decoration System

With the further development of household electricity supplier,the user's demand for personalized recommendation is growing and a variety of recommendation algorithms began to apply.In the traditional clustering recommendation algorithm,each user or item can only belong to one category,so it is difficult to describe the characteristics of each of the objects involved in clustering.The memory based collaborative filtering recommendation algorithm and the improved collaborative filtering recommendation algorithm based on clustering are not accurate enough to solve the problem of sparse data.Aiming at these shortcomings,this paper proposes a hybrid recommendation algorithm LPHRA.The proposed algorithm improves the prediction accuracy and solves the application limitations of data sparsity.The main results of this paper are as follows:1 aiming at the deficiency of LSPM in the category recommendation,this paper puts forward the influence factor and improves the key algorithm EM algorithm.The experimental results show that the influence factors can improve the accuracy of prediction.The EM algorithm allows the user and the attributes of the items to belong to more than one hidden class.LSPM can give the new users an accurate recommendation.2 LPHRA recommends for a specific category of goods by PVH measure.In a certain category of goods,select some good PVH properties the items ane add these into recommended list.In this way,user can get personalized recommendations based on the user's browsing habits.The experimental results show that the proposed hybrid algorithm is more accurate than the traditional collaborative filtering recommendation algorithm and the prediction results are closer to the user preference.3 based on the hybrid recommendation algorithm LPHRA,a home decoration system is designed and implemented.After the system test,the system Home Furnishing decoration based on the LPHRA algorithm is better than the traditional collaborative filtering algorithm and improved clustering collaborative filtering algorithm in the user experience and the prediction accuracy.