Relevant Spatial Data Aggregation Method Based on Clustering Method
Abstract
In the large data age, the number of Internet users is explosively growing, and the spatial network size of social networks, e-commerce trading networks and so on is developing rapidly. It is very important to accurately and effectively cluster spatial data in the collaborative structure of the user points of interest recommendation and hotspot broadcast. A new clustering algorithm for collaborative spatial data clustering is proposed, relevant spatial data aggregation method based on cluster method, shortened as RSACM, The time complexity is O (nlog2 (n)), the algorithm is based on clustering and iterative method to apply the new method of updating the nodes and edges, and use the balanced binary tree method to index the clustering increment. Compared with the traditional overlapping node clustering algorithm, the frequency of each node analysis is greatly reduced, and the recognition accuracy can be obtained at lower algorithm running time. The test results of the algorithm on the spatial data set show that the RSACM algorithm can effectively gather the network cooperative correlation, and the community recognition accuracy is higher. On the large scale LFR reference data set, the cooperative clustering standard of mutual information the highest can reach 0.97, the average value of the clustering index F-score is above 0.91, and the running time under the spatial data is superior to the traditional algorithm.