A New Method of K-Means Clustering Algorithm with Events Based on Variable Time Granularity

According to the characteristics of Weibo event, this paper analyzes the advantages and disadvantages of the traditional K-means algorithm, and proposes the K-means clustering algorithm of events based on variable time granularity. The experiments show that the improved algorithm is more suitable for clustering analysis of Weibo event, improves the efficiency of clustering algorithm, and solves the initial cluster centers sensitive issue, compared with the traditional K-means algorithm.