Applied Research On Home Decoration Field Of Collaborative Filtering Algorithm

Recently,with the advancing development of the e-commerce business the information on the internet is becoming more and more hard for users to locate the useful information.We have limited time and energy,we can locate the target when we know the information about the products.In this situation,the recommend system arises at the historic moment.The recommend system is similar to the people who recommend products in our real life.The economic scale of the home decoration industry almost got 4000 billions yuan,On the one hand,the research on home decoration scheme is very few,on the other side,most of the recommendation system only consider the history factors while neglect the characters of users and items,this article provides a useful collaborative filtering algorithm of home decoration for customers according to learning about the behaviors and characters of users and the features of item.The main work of this paper is as follows:1)I analyze the research and applications of the collaborative filtering recommendation system by reading lots of literature and document,knowing that the applications on home decoration scheme is very few,so i want to take this theory into home decoration scheme.2)I propose a collaborative filtering algorithm that combines the user ratings similarity and the user features similarity to select user nearestneighbors,then put the result into next step.This algorithm can alleviate the data sparsity.3)I propose a collaborative filtering algorithm that combines the item ratings similarity and the item features similarity to select item nearest neighbors and improve the slope one algorithm.4)According to my work experiences in X Internet Home Decoration co.LTD,I analyze the dates of the features and behaviors of customers,in order to verify the useful of the new algorithm,the dates is divided into two parts,one is training set,one is test set.