Feature Extraction of Network Temporal and Spatial Distribution Based on Data Stream Clustering

Spatio-temporal data contains a lot of knowledge information, including patterns and characteristics, the relationship between data and data and their characteristics, etc. How to extract these characteristics from these spatio-temporal data has become the focus of research. To this end, a method of network spatiotemporal distribution feature extraction based on data stream clustering is studied. This method first analyzes the relevant theories of the clustering algorithm, and then uses the DBSCAN algorithm in the clustering algorithm to extract the characteristics of network temporal and spatial distribution. The results show that: compared with the traditional extraction method, the extraction quality of this method is higher, reaching the goal of this paper.