A Recommendation Algorithm Based On User Experience And Long Tail Theory

In the era of sudden increase in information and data,the network users face "inform ation is difficult but valuable information is difficult to troubleshoot and screen" the dilem ma.Filter technology is the most successful measure in the information recommendation s ystem.It was born in the nineties of the 20 th century.At present,because the network users provide reliability recommendation,saving time and convenient experience has become the focus of research.It can quickly and proactively estimate the properties of users and the interests of the direction,push may need but the users is difficult to filter,to determine the relevant information resources.But the data information to the traditional recommendation algorithm has brought new problems,large or sparse data makes the similarity meter is not accurate,the user's satisfaction with the recommendation information decreased,cold start and other issues have been generated.Many scholars have begun to study and improve traditional algorithms for these problems.Based on the traditional recommendation,this paper proposes a recommendation algorithm based on user experience and long tail theory.Firstly,the potential impact of user experience on users is expounded,the formulas of traditional algorithms are briefly explained,and the calculation formula of Pearson's similarity is further calculated.This paper focuses on the evaluation of the behavior of the project and the proportion of the user's historical behavior score in the total project,focusing on the "user score matrix and the project category matrix",by virtue of the long tail Theory to balance the recommended probability of popular and unpopular projects.The experimental results show that the algorithm has some improvement.