基于云计算的社交网络并行推荐算法

Parallel Recommendation Algorithms for Social Networks Based on Cloud Computing

云计算平台的诞生,推动了推荐系统的发展。传统社会网络并行算法在大数据处理方面存在不足,利用签到记录时,不能充分利用签到信息所隐含的偏好、位置和社交网络信息,从而造成准确率低的问题。基于此,提出了基于云计算的社交网络并行推荐算法,Debogone算法设计。通过特征提取算法实现Debogone算法设计,以用户历史偏好为基础,综合考虑用户社交关系,以用户的活动范围为约束点,实现用户兴趣点的推荐。通过实验对比,证明了Debog one算法设计准确率高,稳定性高,具有推广意义。

The birth of cloud computing platform promotes the development of recommendation system.The traditional parallel algorithm of social network has some shortcomings in large data processing.When using check-in records,it can not make full use of the preferences,locations and social network information implied by check-in information,which causes the problem of low accuracy.Based on this,a parallel recommendation algorithm for social network based on cloud computing,Debog one algorithm design,is proposed.Debog one algorithm is designed by feature extraction algorithm.Based on user's historical preferences,user's social relationship is considered comprehensively,and user's activity range is the constraint point to realize the recommendation of user's interest point.The experimental comparison proves that the Debog one algorithm has high design accuracy and stability,and it has the significance of generalization.