基于改进K-均值算法的图书馆图书个性化推荐技术研究

Research on personalized recommendation technology of library books based on improved K-means algorithm

为提升图书馆服务质量,对图书个性化推荐技术进行研究。对传统K-均值算法存在的聚类数目选择与初始聚类中心确定困难问题,设定聚类数目范围,通过迭代自适应确定聚类数目,同时基于密度来确定初始聚类中心,实现对算法的改进。将改进K-均值聚类算法应用于高校图书馆图书个性化推荐中,和传统K-均值聚类算法对比结果表明:当迭代步数为10时,改进K均值算法的挖掘精度相对于传统K均值算法的挖掘精度提高了11.0%;当迭代次数增加到20时,两种算法的挖掘精度相差仅为1.6%,但改进K均值算法所用时间减少了92.1%,迭代效率大大提升。这对提升高校图书馆服务水平具有一定的参考价值。

In order to improve the service quality of the library,the personalized recommendation technology of books is studied.For the difficult problem of selecting the number of clusters and determining the initial cluster center in the traditional K-means algorithm,set the range of the number of clusters,determine the number of clusters adaptively through iteration,and determine the initial cluster center based on density to improve the algorithm.The improved K-means clustering algorithm is applied to personalized recommendation of books in university libraries.The comparison results with traditional K-means clustering algorithms show that when the number of iteration steps is 10,the mining accuracy of the improved K-means algorithm is 11.0%higher than that of the traditional K-means algorithm;When the number of iterations increases to 20,the difference in mining accuracy between the two algorithms is only 1.6%,but the improved K-means algorithm reduces the time by 92.1%,greatly improving the iteration efficiency.This has certain reference value for improving the service level of university libraries.