Research On Algorithms Of Personalized Recommendation System

The development of computer technology makes our society into the information age.The information age has made great changes in our lives.We can find the information we need in the Internet anytime and anywhere.The information age brings us convenience,but also brings some problems,that is the so-called "information overload" problem."Information overload" is one of the negative effects in the process of informationization.It means that in the process of information construction,due to the exponential growth of information,resulting in a large number of redundant information in the network.Itcan't was fully utilized by people.In order to solve this problem,the researchers put forward a lot of methods,one of the most representative is the recommended system.The recommendation system scientifically calculates,processes and analyzes the user's historical data and behavior information,establishes the user's interest model,and recommends the user's favorite content through the interest model.Although the recommendation system can effectively solve the "information overload",but also inevitably face many problems(such as cold start,recommended accuracy and user interest time changes,etc.).Therefore,this paper mainly studies how to improve the performance of the recommended system,solve the cold start of the recommended system and user interest time-varying problem.This paper analyzes the influence of the overall behavior of the user on the recommendation system in the network activity,and puts forward the concept of the tag active cycle.The tag active cycle can reflect the influence of the user's behavior on the recommendation system well.At the same time,it analyzes the influence of the user's tagging time on the overall recommendation,and proposes the tag time weighting factor.Combined with the recommended technical characteristics of the network structure,applying time weighting factor to improve the network structure recommendation algorithm.a new personalized recommendation algorithm based on time weight is proposed.And the algorithm is compared with some classical algorithms.The results show that the algorithm can obtain satisfactory results in Delicious and Movielens datasets.The algorithm effectively improves the accuracy and diversity of the recommendation system.In the further experiment,it is found that the time-weighted personalized recommendation algorithm is better in the two data sets when the weight of the resource object is smaller.The results also prove that the algorithm can solve the problem of "cold start" well.