Research On Minimal Set-covering Model-based Data Placement For Online Social Network

With the development of the Internet and online social networking applications,billions of users are active in various social network platforms and generate massive amounts of data.As the key technology of social networking backstage support system,data placement for online social network is a very worthy research direction in social network data management.Facing with massive social network data,how to make resource scheduling scheme more reasonable is the primary research goal of data placement for online social network.However,there are a lot of redundant copies in the existing data placement scheme,therefore,this thesis transforms the problem of data placement for online social network into the problem of minimum set-covering problem,focusing on the optimization of copy-efficiency.In this thesis,we apply the minimum set-covering model to the online social network data placement problem for the first time by constructing a correlation matrix,and the following two aspects of online social network static and dynamic communities are studied.Firstly,a new heuristic data placement strategy based on minimum set-covering model(LDS)is proposed to solve the problem of huge system overhead and slow system response in the static community of online social network This strategy optimizes operating costs and improves system performance by constructing a Latency-constrained matrix and performing aggregate operations to output a minimum set of copies that meet the quality of service,thereby reducing the number of redundant copies and improving the copy-efficiency.Then for the resource allocation problem of online social network dynamic community,this thesis proposes a dynamic data placement strategy based on the minimum set-covering model(LDS),which solves the resource scheduling problem under the scenarios of increasing or decreasing social network users and friendships to optimize operating costs and system performance of the dynamic community.This thesis proposes a data placement strategy based on minimum set-covering model,and conducts in-depth research on static community and dynamic communityof online social network.Finally,experiments prove that the strategy of this thesis can effectively improve the copy-efficiency,optimize the operating cost,improve the system performance and ensure the quality of service,which has certain theoretical and practical significance for reducing energy consumption and promoting energy conservation.