摘要
为进一步解决基于用户的协作过滤技术的扩展性问题,利用基因表达式编程(GEP)的并行性优势,与已有的串行聚类DBSCAN算法进行融合,使得串行程序并行化,提出了一种GEP-DBSCAN协作过滤聚类算法来寻找最近邻居,改进基于密度的协作过滤方法,实验证明了算法的有效性以及提高了时间效率.
To address the expansibility problem of collaborative filtering technology based on users,the GEP-DBSCAN algorithm for collaborative filtering clustering was proposed.It is key to fuse the parallelism of gene expression programming and the advantages of the DBSCAN algorithm.The new algorithm makes serial programs parallelized and can be used to find the nearest neighbors.It improves collaborative filtering method based on the density.The experimental results show that the GEP-DBSCAN algorithm is effective and can increase time efficiency.
出处
《中南民族大学学报（自然科学版）》
CAS
北大核心
2017年第4期112-115,共4页
Journal of South-Central University for Nationalities：Natural Science Edition
基金
国家自然科学基金资助项目(61262028)
关键词
聚类算法
基因表达式编程
协作过滤
cluster algorithm
gene expression programming
collaborative filtering
