项目名称: 支持大型社交网络的云存储系统
项目编号: No.61202047
项目类型: 青年科学基金项目
立项/批准年度: 2013
项目学科: 计算机科学学科
项目作者: 伍赛
作者单位: 浙江大学
项目金额: 23万元
中文摘要: 社交网络的用户量和数据都在急速增长,在我国有超过1.6亿人在使用社交网络。社交网络数据处理系统必须能够处理海量数据,能够支持高并发查询,能够对数据进行有效的分析。传统的数据库系统已经无法支持新型的社交网络应用,成为了制约社交网络发展的一大障碍。针对这一问题,本研究将云计算技术和社交网络相结合,通过设计新型的云数据处理平台来利用云计算的无限计算资源为社交网络服务。为了将社交网络存储在云平台上,我们将研究如何把社交网络数据转化为图模式,然后按照分而治之的策略将整个社交网络图切分为若干子图,每个子图都被分散到不同的云计算节点来进行处理。在这一策略中,本研究需要解决一系列的科研难题,比如:图划分问题、分布式索引问题、并发查询引擎问题以及社交网络实时分析等。每一个子项都涉及到目前国内外非常前沿的科研课题。本项目将研发一个原型系统部署在真实的云环境中。具体的研究成果,将发表在国内外知名刊物和会议中。
中文关键词: 社交网络;分布式索引;列存储;K匿名;个性化推荐
英文摘要: The number of users increases sharply in the social network. More than 160 million people in China start using the social network to share their photos and videos. Such a large scale users and data pose new challenges for the data management system. The conventional database system fails to deliver a sastisfied performance for the social networks. Therefore, in this project, we propose to exploit the Cloud computing techniques to support the applications of social networks. The intuition is to model the social network as a large graph and apply the graph partitioning techniques to split the graph into many sub-graphs. Each sub-graph will be dissemnated to a specific Cloud node, which responds for handling data of users in the sub-graph. Then, we can use the Cloud algorithms to handle the social network problems. The above strategy involves a set of research topic, such as graph partitioning, distributed indexing, concurrent query engine and social network real-time analysis. All the topics are very challenging and will attract the interest of researchers from different communities. In this project, we plan to deliver a arch-type system, which can be deployed on the real Cloud environment, serving current social networks. Our research work will be published in a few top conferences and journals.
英文关键词: Social Network;Distributed Index;Column-oriented Store;K Anonymity;Personal Recommendation