Abstract: For very large databases，such as spatial database and multimedia database，the traditional clustering algorithms are of weaknesses in effectiveness and scalability.According to the notion of clustering feature of BIRCH，a dynamic and incremental clustering algorithm is designed and implemented，which solves the problems of effectiveness，space and time complexities of clustering algorithms for large databases.Theoretic analysis and experimental results demonstrate that the dynamic and incremental clustering algorithm can not only handle large databases，but also has good scalability.

摘要： 对于大型数据库，如空间数据库和多媒体数据库，传统聚类算法的有效性和可扩展性受到限制。通过动态增量的方法，在基于密度和自适应密度可达聚类算法的基础上，根据BIRCH算法中聚类特征的概念，利用簇特征设计与实现了一种新的动态增量聚类算法，解决了大型数据库聚类的有效性以及空间和时间复杂度问题。理论分析和实验结果证明该算法能够有效地处理大型数据库，使聚类算法具有良好的可扩展性。