Design And Implementation Of Book Recommendation System Based On Hybrid Strategy

Nowadays,we are in the era of information overload,users can not quickly find the items they are interested in from the numerous information.At the same time,businesses are also facing a problem,they are difficult to put goods to the right users.In this case,recommendation system came into being,it can not only bring convenience to users,but also bring huge profits to enterprises.Many companies are deeply aware of the potential value of data,so they establish their own recommendation system platform to enhance the user’s stickiness,so as to increase the sales revenue of enterprises.This paper presents a hybrid recommendation model based on stacking integration strategy to improve the accuracy of recommendation results.The main work of this paper is as follows:Firstly,the common algorithms of recommendation system include domain based,model-based,content-based,etc.This paper introduces the basic principles,advantages and disadvantages of each algorithm mechanism.Secondly,due to the good generalization ability of ensemble learning,it is widely used in many fields.This paper combines ensemble learning with recommendation algorithm,and proposes a hybrid recommendation algorithm based on ensemble learning.The idea is:using stacking integration strategy,taking item based collaborative filtering recommendation algorithm and LFM algorithm as the base model,The basic model is trained and verified by 50%cross validation,and the results of each validation set are combined as the input of the meta learner.The xgboost algorithm is used as a meta model for prediction,and the prediction result is the final recommendation result.Third,this paper uses crawler technology for data acquisition and preprocessing.Recall,accuracy,F1 and RMSE are used to compare the performance of LFM,project-based and hybrid models.Experimental results show that xgboost hybrid model has higher accuracy than LFM and project-based recommendation.Fourth,according to the hybrid recommendation model based on integrated learning,the book recommendation system is designed,including the system architecture design,database design and the implementation of the web end.