Social network community partitioning method and server

Abstract

The embodiment of the invention discloses a social network community partitioning method. The social network community partitioning method comprises the following steps: constructing a social networkmodel based on social relationship in a social network, and giving weights to mutually connected edges between nodes in the social network model, wherein each node carries a tag; iterating and updating the tags of all the nodes in the social network model according to the tag of each node and the weights of the edges of each node; and when number of times of iteration and update is equal to a preset threshold value, partitioning user identifications corresponding to the nodes carrying the same tag in the updated social network model into one community. The embodiment of the invention also discloses a server. By adopting the social network community partitioning method and the server which are disclosed by the embodiment of the invention, stability of the tag of the node updated by the server can be improved, so that stability of the server during community partitioning is improved.

Description

technical field [0001] The invention relates to the field of social networks, in particular to a method for dividing social network communities and a server. Background technique [0002] With the rapid development of the Internet, the social network formed by users in the network is no longer just a relationship between a single user and a single user, but a one-to-many and many-to-many relationship. At this time, users in the social network can be divided into communities according to their social relations in the social network, so that users can be managed and maintained more effectively, and targeted services can be pushed for the communities. [0003] At present, community division of social networks mostly adopts LPA (Label Propagation Algorithm) algorithm for community mining division. However, when using this algorithm for label update, it only obtains the optimal label based on the label of the neighbor node of the current node for label update. Makes it possible...

AI Technical Summary

Problems solved by technology

However, when using this algorithm for label update, it only obtains the optimal label based on the label of the neighbor node of the current node for label update. Makes it possible to have multiple optimal labels

When there are multiple optimal labels, the algorithm will just randomly select one of the optimal labels for label update. Due to the randomness in the label selection process, the algorithm is unstable. Therefore, using this The same result cannot be obtained when the algorithm divides multiple times, so the stability of the divided community cannot be guaranteed

Images