Spatial Selection of Sparse Pivots for Similarity Search in Metric Spaces
Spatial Selection of Sparse Pivots for Similarity Search in Metric Spaces
复制标题
度量空间中相似性搜索的稀疏枢轴的空间选择
DOI：
10.1007/978-3-540-69507-3_37
复制
发表时间：
2007
期刊：
影响因子：
--
通讯作者：
N. Brisaboa
中科院分区：
文献类型：
--
作者：
Oscar Pedreira;N. Brisaboa
相似性搜索是处理非结构化数据源的应用程序的必要操作。在本文中，我们提出了一种基于主元的方法，不仅可以在不预先指定主元数量的情况下获得良好的主元选择，而且还可以深入了解度量空间的复杂性。稀疏空间选择（SSS）适应度量空间的维数，是动态的，适合辅助内存存储。在本文中，我们提供了实验结果，证实了该方法在多个度量空间上的优点。此外，我们还解释了如何轻松并行化 SSS。最后，在本文中，我们概念化了嵌套度量空间，并证明，在某些应用领域，对象可以分组在具有不同关联度量空间的不同簇中，所有这些都嵌套到解释簇之间距离的通用度量空间中。
Similarity search is a necessary operation for applications dealing with unstructured data sources. In this paper we present a pivot-based method useful, not only to obtain a good pivot selection without specifying in advance the number of pivots, but also to obtain an insight in the complexity of the metric space. Sparse Spatial Selection (SSS) adapts itself to the dimensionality of the metric space, is dynamic, and it is suitable for secondary memory storage. In this paper we provide experimental results that confirm the advantages of the method with several metric spaces. Moreover, we explain how SSS can be easily parallelized. Finally, in this paper we conceptualize Nested Metric Spaces, and we prove that, in some applications areas, objects can be grouped in different clusters with different associated metric spaces, all of them nested into the general metric space that explains the distances among clusters.