摘要
针对人工智能领域图聚类数据分析与处理能力无法适应于日益复杂的分布式集群环境等问题,设计出一种基于并行计算的高效率图聚类信息处理方案。通过对Minhash算法以MapReduce架构理论进行改进,使其实现对数据的并行化分析处理,以确保其能够在日益复杂的分布式集群计算环境下高效处理图聚类数据信息。通过相关实验表明,该方案不仅可行,而且能够对图聚类数据信息进行快速稀疏化处理,具有一定的高效性。
In order to solve the problem that the analysing and processing abilities of graph clustering data in artificial intelligence field can＇ t adapt to the increasingly complex distributed clnster environment, we design a parallel computing-based efficient graph clustering information processing scheme. In this scheme, the Minhash algorithm is improved based on MapReduce framework theory to enable it to achieve the paralleled analyses and processing on the data, so as to guarantee it being abIe to efficiently process graph clustering data information in increasingly complex distributed cluster environment. It is indicated by related experiment that this scheme is more than feasible, it can also quickly carry out sparseness processing on graph clustering data information, and has certain high efficiency.
出处
《计算机应用与软件》
CSCD
2016年第2期217-222,共6页
Computer Applications and Software
基金
国家自然科学基金创新研究群体科学基金项目(51021004)
