改进的Louvain算法及其在推荐领域的研究

Optimization of Louvain algorithm and its application in personalized recommendation

文中在深入研究社区发现算法、个性化推荐技术和Spark集群实现技术的基础上,提出了基于叶子社区的社区发现算法和对应的个性化推荐设计方案。采用Scala语言,结合Spark Graph X图计算技术在Spark集群上实现该方案。与传统推荐技术相比,基于社区发现的个性化推荐方法在推荐效率以及正确度方面都得到大幅度提升。

For the community discovery algorithm,personalized recommendation technology and Spark cluster implementation technology,this paper proposes a community discovery algorithm based on leaf community and the corresponding personalized recommendation design scheme. Scala language,combined with Spark Graph X graph computing technology in the Spark cluster to achieve the program.Compared with the traditional recommendation technology,the personalized recommendation method based on community discovery greatly improves in the recommendation efficiency and accuracy.