【摘要】 <正> 1 引言近年来,高维数据库的应用得到快速的发展,如海量的多媒体数据库、大规模的文本数据以及生物信息学中庞大的DNA数据库等,这些信息一般使用特征抽取等方法映射为高维数据,然后通过计算这些高维数据之间距离实现相似性查询。例如,对于图像数据,往往采用颜色直方图来表征一幅图像,当需要从数据集查找与给定图像相似的图像时,通过计算更多还原

【Abstract】 Fast searches and query operations in high dimensional databases require efficient index structures. Among a variety of index structures, the index structures in metric spaces are very useful. They can be used in an extensive field, such as searching for protein molecular chains with certain sequences in Computational Biology and matching a given strings fuzzily in Text Retrieval. In this paper, the features of index structures in metric spaces are analyzed and subsequently a further classification is given to these index structures. Finally, some representative index structures are introduced in detail.更多还原