蚁群优化与模糊聚类结合的文本聚类研究

Research on document clustering based on ant colony combined with Fuzzy C-means

针对模糊文本聚类算法(FCM)对输入顺序以及初始点敏感的问题,提出了一种使用蚁群优化的模糊聚类算法(FACA)。该算法采用蚁群聚类算法(ACA)找到聚类的初始中心点,以解决模糊聚类的输入顺序以及初始点敏感等问题。模糊文本聚类算法的线性复杂度使其更便于在计算机实现。与经典的基本模糊聚类以及蚁群聚类在真实数据集上仿真相比较,结果表明经蚁群优化过的模糊聚类算法(FACA)效果更有效,更适合应用于大型的数据集。

Focusing on the problem that the Fuzzy C-Means clustering algorithm is sensitive to initial centers and input order,a document clustering algorithm combined with ant colony clustering and Fuzzy C-Means is proposed.The algorithm takes advantages of ant colony clustering algorithm to find the initial centers,then uses Fuzzy C-Means to get the accurate result.Experimental results show the good performance of the hybrid document clustering algorithm,and it is better for the large-sized dataset.