Microblog Hotspot Discovery Method Based on Improved K-Means Algorithm

Abstract

The K-means algorithm is one of the most frequently used clustering algorithms in hot topic discovery. However, due to its shortcomings such as the number of clusters K value and easy to fall into local optimum, the clustering accuracy is not high, which directly affects the quality of hotspot discovery. This paper proposes an improved K-means algorithm to achieve fast clustering of microblog texts. Combining the high-frequency words and similarities of the microblog texts to perform single-pass clustering, the K number of clusters and the initial clustering center are obtained, which solves the problem that the K-means algorithm is too sensitive to the K value and the initial center. Through experimental comparison and analysis, it makes up for the shortcomings of K-means algorithm, and effectively improves the efficiency and accuracy of clustering. Applying it to the hot topic discovery model, the effectiveness of the hot spot discovery model based on the improved K-means algorithm is verified by experiments, and it has a high accuracy.