Sequential network and sequential data polymorphic clustering method

Abstract

The invention provides the concept of polymorphic clustering and discloses a sequential network and sequential data polymorphic clustering analysis method. Polymorphic clustering is a polymorphic clustering analysis method with multiple standards as similarity measurement indexes and used for a sequential network and sequential data. The method includes the steps that first, if processed objects are sequential data, the sequential data are only processed into a special form, and if the processed objects are the sequential network, the sequential network is mapped into sequential data in a special form with a spectrum mapping method; second, polymorphic vectors of the sequential data are established; finally, an improved synchronous clustering method is adopted for clustering the polymorphic vectors to obtain the polymorphic clustering result. The method allows people to observe a community structure of the sequential network or the sequential data from different angles, and therefore a more comprehensive analysis result can be obtained.

Description

technical field [0001] The invention belongs to the technical field of data mining and graph mining, and more specifically relates to a time series network and a polymorphic clustering method of time series data. Background technique [0002] In order to identify the community structure in time-series networks and time-series data, popular clustering methods in static network and static data analysis can no longer meet the requirements. In 2006, Chakrabarti et al. proposed the concept of evolutionary clustering. They clustered data over time, and proposed evolutionary versions for two commonly used data clustering methods, k-means and synthetic clustering algorithm (Agglomerative Hierarchical Clustering). [0003] However, in the existing methods, the clustering of complex networks uses the connection structure between nodes as the clustering standard, and through this clustering, the nodes that are closely connected with each other are divided into the same cluster. Most s...

AI Technical Summary

Problems solved by technology

Most spatial clustering methods only use the spatial distance of point objects as the similarity measure for clustering, and only consider the spatial proximity characteristics of point objects, while ignoring non-spatial attribute features.

Images