Social user data optimization placing method based on hypergraph segmentation

Abstract

The invention relates to a social user data optimization placing method based on hypergraph segmentation, and the method comprises the three steps: constructing a social network model based on a hypergraph, carrying out hypergraph dividing and carrying out the hypergraph-to-subgraph mapping. The method can achieve the accurate description of multi-user interaction behaviors in a social network through the hyperedge in the hypergraph, achieves the placing of the user data according to the interaction behaviors of a user, can greatly reduce the communication quantity generated in an interaction process of the user, can maintain the load balance between servers while optimizing the data placing. The method can optimize the placing of the user data in each server according to the interaction behavior of a social network user, maintains the load balance between the servers, and reduces the communication quantity in a data center.

Description

technical field [0001] The present invention relates to the fields of social network, big data and distributed storage, in particular to a method for optimal placement of social user data based on hypergraph segmentation in a social network storage system. Background technique [0002] Nowadays, Online Social Networks (OSNs for short) such as Facebook, Twitter and WeChat have become one of the most popular Internet applications. Taking Facebook as an example, its monthly active users reached 1.59 billion at the beginning of 2016, equivalent to nearly a quarter of the world's population using Facebook as a social tool to communicate on the Internet. The reason why online social networks are so popular is that apart from the popularity of the Internet, it is more important that they can project people's social relationships and activities into cyberspace, making communication more convenient, fast and efficient, especially for users who are geographically far apart. [0003]...

AI Technical Summary

Problems solved by technology

The existing key value storage system adopts the hash method, and the user data storage location is randomly determined by the hash value. The hash enables the data of any user and his friends to be stored in multiple servers. User interaction requires communication between servers. The number of users increases rapidly, and the traffic in the data center increases sharply, and the huge internal traffic will cause network congestion, increase data access delay, and reduce system scalability

[0007] Aiming at the high traffic brought by the hash method in the existing key value storage system, the present invention provides a social user data optimization placement method based on hypergraph segmentation, modeling the social network user interaction relationship as a hypergraph, and using the hypergraph Segmentation solves the user data placement problem to reduce traffic in the data center while maintaining load balance between servers

Images